Darwin himself, as we shall see later, once expressed such a
view.
On what might we possibly base such a broader
theory? In Chapter 3, I suggested that there was a close
relationship between information and holarchical development. A
cell contains more information than an atom, and a multicellular
organism more than a cell. Furthermore, multicellular stages
have more information than single cells, and societies of
organisms more than their individuals. This suggests that
evolution of the holarchy might be considered a process of
increasing information, an idea that has attracted several other
scientists (Sheldrake, 1981, 1989; Chaitin, 1999; Davies, 1999;
Loewenstein 1999).
In this chapter, I will attempt to build a
basis for integrating the concept of evolution with information.
In particular, I will argue that we can use information--in
theory, if not quite yet in practice--to quantitate the
evolutionary process. That is, information is the yardstick by
which we can measure how much evolution has taken place. If this
can be established, understanding evolution becomes a matter of
understanding how information is created, stored and transmitted
by living things.
The Relationship of Genetic Information to
Evolution
At the beginning of this chapter, I pointed
out that evolution of the holarchy involves three kinds of
processes: diversification, transformation and transcendence.
Diversification produces new holons on the same stage of
existence; transformation produces new stages on the same level;
and transcendence produces a new level of existence. At first
glance, these three processes may seem quite distinct. However,
all of them can interact, which can make it difficult to define
a particular evolutionary process. For example, the evolution of
different organisms might seem to be a diversification process,
but it clearly involved transformation at some points as well.
The emergence of vertebrate organisms was to a great extent a
transformative process, as was the evolution of our own species
from other primates. The transformations involved the creation
of higher stages of biological existence--more complex brains,
for example--and in the case of our own species, higher stages
of mental existence also, that is, social organizations.
At the outset, then, we need a more precise
definition of diversification, a way to draw the line more
clearly between this kind of process and transformative ones.
Suppose, as a preliminary attempt to address this issue, we
define diversification as any evolutionary process which does
not involve an increase in information. More specifically,
let's say diversification results in a change in the quality
of information, but not in the quantity of information.
The change in the quality of information is reflected in the
evolution of a new form of life; the lack of change in amount of
information is reflected in the fact that the new holon is on
the same stage of existence as that from which it evolved.
Consider a typical, clear-cut example of
diversification, one in which we can agree no transformation has
occurred: the evolution of a new species of organism from
another species of the same family or genus. The new species has
certain phenotypic features that distinguish it from the former
species. For example, a new species of bird might have a larger
bill, a different color plumage, or different nesting habits
from that of the related species from which it evolved. These
new features, I argue, represent differences in the quality of
information the organism expresses, but not in its quantity.
Each species of bird contains approximately the same quantity of
information.
This definition of diversification obviously
will not be very useful, of course, unless we can define, and
hopefully quantitate, information itself. What do I mean when I
say an organism expresses a certain quantity of information,
which is no greater than that expressed by another organism of
the same species? In the conventional scientific view, the
information specifying any organism is contained in its genome.
With this in mind, we could define diversification as a process
by which the quality of genetic information changes, but not
it's quantity. Though we have no completely reliable method of
determining the quantity of genetic information any organism
has--a crucial point I will return to later--we can estimate
fairly accurately its number of genes. So let's define the
quantity of information present in any organism--again, in a
preliminary manner--as directly related to the number of its
genes.
Applied to our previous example, this
definition of information seems to be quite suitable and
appropriate. Two closely-related species of birds will indeed
contain about the same amount of genetic information, as
measured by the total number of genes they have. Where they
differ is that a few of these genes are found only in one
species, while a few others are found only in the second
species. Thus one bird may have a gene or several genes that
determine the formation of a particular kind of beak, while the
other species has different genes that determine a different
type of beak.
So far, so good. However, we quickly discover
a problem with this definition of information. It's possible to
identify organisms which most of us would agree are very
different, yet which seem to contain about the same quantity of
genetic information. A classic example is provided by comparing
our own species with that of a non-human primate, such as a
chimpanzee. The evolution of human beings from other primates,
as I noted earlier, was not a simple diversification process. It
involved transformations as well, particularly in the emergence
of our larger brain (Preuss and Kaas 1999). Yet our genome is
very similar to that of a chimp's. We have virtually the same
number of genes (Gearhart and Kirschner 1999) and about 98% of
those genes are the same (Stebbins 1983). While we don't know
the exact number of genes in either humans or chimps, whatever
differences there may be seem highly unlikely to provide a
meaningful explanation of the difference between ourselves and
chimpanzees. That is to say, information, if it's to be related
to holarchical development, can't be equated with any simple
quantity of genetic material.
The similarity of chimpanzee genes to our own
is well-known to scientists, of course. It's simply assumed that
the relatively few genetic differences between the two species
are sufficient to account for the very great differences in
phenotype. For example, it's a very good bet that many uniquely
human genes control the development of our brain, particularly
the cerebral cortex, which is much larger and more complex than
that of other primates (Preuss and Kaas 1999).
This widely held view, however, regardless of
whether it's correct or not, does not address the information
gap--the fact that our genomes appear to contain roughly the
same quantity of information as those of other primates. To
understand why this is so, let's recall the discussion, in
Chapter 3, of Gregory Chaitin's information theory. Chaitin
introduced the critical notion of compressibility. If two
strings or programs are of identical length, the less
compressible one is said to contain more information. By this
definition, the information represented in any organism is
highly compressed. This is why a relatively few genes--roughly
one hundred thousand in the human organism--can specify
literally billions or trillions of cellular interactions. The
highly patterned, repetitive nature of holarchical organization
makes it possible for the genome to express all these
interactions simply by specifying a few cell types and their
rules of interaction.
Chaitin's theory is certainly compatible with
the possibility that two organisms with roughly similar
quantities of genetic information could not only have a very
different structure--i.e., body type--but that one could have a
much larger brain than the other. In this case, the larger brain
would represent a greater "decompression" of the genetic
information; this information would simply specify the growth
and connection of a larger number of cells. The key point,
however, is that the larger human brain could not contain any
more information, in Chaitin's sense, than the smaller
chimpanzee brain. The human brain may have the capacity to form
many more connections among its neurons that the brain of other
primates, but these connections, themselves, do not constitute
more information. Some process still must select particular
connections in order to create more information. "No program
can generate a number more complex than itself," Chaitin tell
us;
"a one pound theory can no more
produce a ten-pound theorem than a one hundred pound
woman can birth a two hundred pound child."
This is fine as an argument against a strict
deterministic role for genes; other factors do interact with
genes. However, it clearly does nothing to explain why a human
brain is so much more developed than a chimpanzee's or a
mouse's. The same program may run differently on
different hardware, but the hardware can't change the amount
of information in the program. Actually, the maternal
environment is not just hardware; it may also add some
information to the developing organism. But the amount that can
be added is nowhere near enough to account for the differences
we are concerned with. All the mother can add is more compressed
information, such as some of her own genes; there is no way she
can add to the embryo the information in her brain. She can't do
this, because as I just emphasized, most of this information
isn't in the genes to begin with. And even if she somehow
could add such information, there is no apparent way the embryo
could make use of it. The "ground plan" of the embryo still
lacks essentially all of the information that will later be
found in the brain.
A second argument against the relevance of
the maternal environment for explaining gene-brain information
differences is that as Cohen and Stewart note, the maternal
benefit, to the extent that it does exist, is for all mammals.
It does not address the issue of why some mammals have
much more information in their brains than others.
Still another means of reconciling genetic
information with evolutionary development may lie in a revision
of Chaitin's definition of information. This definition is based
on an analysis of computer programs, but are we really certain
that we can apply it to living systems? Is it possible that
information in the latter sense is somehow different? I believe
this argument may have merit, but not as a means of reconciling,
or equalizing, the amount of information in genome or brain.
Rather, it seems to me that it would much more likely explain
how such a gap came to be created and maintained in the first
place. I will return to this point later.
In conclusion, I believe the gap between the
amount of information contained in the genome and that expressed
in the organism is real. Therefore, either a) some organisms
have more information in their genomes than they express, or b)
other organisms have more information in their brains than can
be accounted for by their genomes. If we accept a), we must
conclude that the eukaryotic genome reached its evolutionary
culmination--in terms of information content--with the emergence
of mammals. At this point, it contained virtually all the
information needed to specify human beings, even though we did
not evolve until some millions of years later. In a sense, then,
our species has existed--in terms of potential--for a far longer
period of time than our natural history suggests. The central
problem of our evolution--the acquisition of enough information
to specify our larger brain--was solved long before we actually
emerged as a distinct species.
If we take this possibility seriously, then
we must regard the eukaryotic cell as far more evolutionarily
advanced than it's usually considered to be. Indeed, we must
conclude that--by an informational definition--essentially no
further evolution occurred after the emergence of this cell as
it exists in lower mammals. All further evolution, through the
primates and human beings, reflected simply new ways to tap this
pool of information, to express increasingly greater quantities
of it. A mouse, in this view, has just as much potential for
intelligence as we do, but its developmental programs are unable
to make the full use of this potential.
This conclusion, it seems to me, is
completely inconsistent with any "bottom-up" theory of
evolution, that is, one beginning with physical matter. It's
difficult enough--in the view of some scientists, too
difficult--to understand the evolution of cells even if we
accept that every emergent feature provided some new measure of
adaptive fitness, the ability to survive and reproduce better
than its predecessors. But this difficulty is greatly magnified
if we are to believe that the cell actually evolved with a great
quantity of information that was latent, that simply sat
in the nucleus, taking up precious space and energy, and doing
nothing at all. To be sure, an important part of evolutionary
theory is that evolved features that serve one purpose can
sometimes later be put to some other purpose. Evolution may
"discover", so to speak, a new way to use some adaptation, which
evolutionists sometimes call a process of pre-adaptation
(Mayr 1988; Gould 1995). But it's virtually impossible to
imagine large quantities of information in the genome serving
any purpose except to code for proteins.
The precocious evolution of the eukaryotic
cell could become understandable, of course, if we scrapped the
idea of bottom-up evolution, and postulated the existence of an
intelligent creator. Most explanations of of our origins of this
kind assume that this higher form of life either created
everything at once--the fundamentalist view--or simply started
the evolutionary process in motion with the creation of the
physical universe. Neither of these views seems to fit very well
with the idea of advanced eukaryotic cells. Biochemist Michael
Behe, however, who accepts a limited role for evolution, takes a
different creationist view. In a devastating critique of
Darwinism, he argues that the enormously complex and detailed
functional arrangements of proteins in the eukaryotic
cell--exemplifed by ciliary motion, blood clotting, and the
formation of antibodies--can't possibly be explained by random
variation and natural selection (Behe 1996). Thus he concludes
that cells were the product of an intelligent designer, and
seems to believe that they were the real starting point of
evolution on earth.
Even some scientists who accept the
established view of evolution have suggested that life on earth
might have been begun at the level of the cell. It could be that
cells evolved elsewhere, and reached earth by chance or even by
design (Hoyle 1981; Davies 1996). This scenario, however, does
not address the problem of how the first cells evolved, and in
Behe's case, we have the problem of the origin of the
intelligent designer. Furthermore, there is substantial evidence
that advanced eukaryotic cells were preceded on earth by cells
with smaller genomes, including the prokaryotes. Thus it does
not seem that such life-by-design arguments--whatever their
other merits--provide an entirely satisfactory solution to the
information gap. In Chapter 12, I will further discuss the
possible role of intelligence in the origins of life.
So let's consider the alternative explanation
for this gap: that primates, and particularly our own species,
have far more information than is accounted for in the genome.
This conclusion surely gives new meaning to the notion of
emergence. Understood in this way, organisms like ourselves have
not simply properties, but quantities, of
something that can't be understood in terms of their components.
Somehow, in the process of evolving, we must have acquired
information from extra-genetic sources. Nor is this just a
problem for evolution. As I pointed out earlier, the same
disconnect between genome and organism also implies an
information gap that has to be crossed every time a new organism
is created. If the information in the genome is not sufficient
to specify the information in the developed organism, where does
that latter come from? We have, it seems, an evolutionary
mystery occurring right here and now. The so-called miracle of
birth really begins to look like one.
Perhaps there is another way of looking at
the problem, though. As I discussed earlier (Chapter 3), there
seems to be a close relationship between energy and information.
We accept that living organisms are capable of accumulating
energy from their environment. As Roger Penrose (1989) has
pointed out, we usually don't so much gain energy as make up for
that we lose. Yet in the process of maturing, a newborn organism
surely does gain energy; its body accumulates far more of it
than it possessed at birth. Is it not possible that a developing
organism can also accumulate information? That is, could a
newborn child, independently of whatever information it gains
from growing up in a complex human society, also gain
information of another kind from its environment? And could
relatively small differences in information at birth result in
much larger changes in information later gained?
But where exactly does this information come
from, and how is it accessed? Several theorists have proposed
the existence of extra-genetic information. In Chapter 11 I will
discuss Rupert Sheldrake's morphic fields, hypothesized
non-manifest forms of existence that can shape evolution,
development, learning and other processes of change by a process
of information transfer (Sheldrake 1981, 1989). It's also
possible to interpret the collapse of the wave function in
quantum physics in terms of the transfer of some unusual kind of
informaton (see Chapter 5). The notion of extra-genetic
information is is obviously a very speculative idea, and I will
put off further discussion of it until later.
So whichever way we wish to interpret the
information gap, we are led to some rather remarkable
conclusions. I believe that eventually it will be possible to
measure the information in both the genome and the brain of
organisms in an accurate and meaningful way, and at this point
we may be able to distinguish which of these possible
interpretations of the gap is correct. The second
interpretation, however--that there is more information in the
brain of higher vertebrates than in the genome--seems far more
likely to me. The brain, after all, has billions of neurons,
each of which may have dozens, hundreds or even thousands of
connections with other cells. This is many orders of magnitude
above the most liberal estimates of the number of genes.
The Brain, Information and Evolution
From the preceding discussion, we can
conclude that it's not possible to use genetic information to
quantitate evolutionary changes in a consistent manner. There is
a correlation between the two in some cases. Organisms that are
highly similar to each other do have similar amounts of genetic
information; and some major evolutionary transitions, as
represented in going from bacteria to eukaryotic cells to
multicellular organisms, are accompanied by large increases in
genetic information. But many large evolutionary changes are not
reflected in significant changes in genetic information.
Perhaps this should not be surprising. The
genome is the informational holon for the cell, and we would
expect that transformative changes in single cells, such as
occurred during the evolution of prokaryotes to eukaryotes,
would be accompanied by an increase in genetic information. But
with the evolution of organisms, a new kind of informational
holon emerged, the brain. In Chapters 3 and 4 I discussed the
ways in which the brain is analogous to the genome, playing much
the same role on its level of existence as the genome does on
its level. Thus the brain regulates the function of biological
stages in the organism, as the genome regulates the function of
physical stages in the cell; and the brain specifies the
development of higher stages of mental existence, as the genome
specifies the development of higher stages of biological
existence.
Given these analogies, we might ask whether
we can use information in the brain to quantitate evolution.
There is clearly a strong correlation between development of the
brain, on the one hand, and evolution, on the other. The first
multicellular organisms really had no brains, in the sense the
term is usually used, but simple nerve nets, distributed
diffusely throughout the body. In higher invertebrates, such as
crustacea, molluscs and insects, nerve cells are concentrated in
groups called ganglia, some of which are found in the anterior
portion of the body and function as a primitive brain (Bullock
and Horridge 1965). In vertebrates, centralization of the
nervous system becomes the primary feature, with a clearly
defined brain and spinal cord. In the higher vertebrates, the
brain becomes increasingly larger (Preuss and Kaas 1999).
There is, to be sure, no precise way to
correlate brain development with evolution. Currently,
scientists know even less about how to measure information in
the brain than in the genome. The task is further complicated by
the fact that as some species develop--particularly ourselves,
of course--the amount of information in the brain seems to
increase dramatically, through learning and other kinds of
experience. The ability to learn is certainly relevant to
evolution, but if we are to use information as a measure of
evolutionary development, we want to separate these two factors.
We want to be able to quantitate the information present in the
brain at birth, treating as an additional problem how
information in this organ increases later on.
I see no reason why in principle science
should not eventually be able to do this. As I discussed in
Chapter 4, we are beginning to understand a great deal about the
brain correlates of mental functions like thinking, learning and
perception. Such studies suggest that it may be possible to
define information in the brain in terms of the numbers of
particular kinds of synaptic connections. Even if the way the
brain stores information turns out to be much more complicated
than this--for example, if information is distributed throughout
the brain, by some quantum (see Chapter 5) or classical
process--it may still be possible to quantitate it. If we can do
this we will have a way of measuring evolutionary change. We
will be able to say that the differences in information content
of the brains of two organisms reflects the differences in their
evolutionary development.
An evolutionary biologist might argue that
there are many kinds of evolutionary changes besides those
relating to the development of the brain. Many other kinds of
organs have of course evolved. But my larger point is that the
evolution of all other organs is ultimately related to that of
the brain. To reiterate the key point: the brain is the
biological analog of the genome, containing all the information
needed to control the biological activities of the organism.
From this it follows that major evolutionary changes in all
other organs or tissues must be accompanied by changes in the
brain. For example, a transformative process involving a change
in the heart, the lungs or the gastrointestinal tract will
invariably be associated with changes in regions of the brain
that control these organs. If, and only if, these changes
involve substantial increases in the amount of information in
the brain, can we define the changes in the other organs--and in
the new organism as a whole--as transformative. Otherwise, the
change is translational.
Even if we define evolution in terms of
information in the brain, however, we soon run into another gap.
Human societies contain far more information than is present in
the human brain. This gap is quite obvious even if we compare
the total information content in society to that in the brain of
a single, mature, educated adult. But the information in the
latter, by virtue of his participation in society, is already
far more than the information present in a human at birth. Where
does all this information come from?
Again, the simple answer, the seemingly
obvious answer--that we are capable of acquiring, transmitting
and storing new information among ourselves--simply does not
address the problem as it has been defined by Chaitin. To
reiterate, because it is such a fundamental point: no program
can generate a program containing more information than itself.
If the amount of information present in the human brain is far
less than that in human society we are again forced to the
conclusion that there is a source of information outside of the
system as we ordinarily understand it, that is, outside of the
physical, biological and mental processes of the organism.
It seems, therefore, that there is an
information gap at every level of existence. There is more
information in the genome than in any individual atom; there is
more information in the brain than in the genome; and there is
more information in our evolving planetary culture than in the
brain. Somehow, evolution must acquire information from
somewhere as it creates a new level of existence. In the next
section, we will try to localize this gap more precisely, see
exactly where the new information comes into play.
Transformation: The Role of Deep and Surface
Structures of Information
To summarize the discussion so far, I'm
trying to define evolution in terms of information. In
diversification processes, exemplified by the evolution of a new
species of organism from a similar species, there is simply a
change in the quality of information, but not in the quantity of
information. In transformation processes, in which a new kind of
organism emerges (a new class or order, for example), there is a
change in the quantity of information. For lower organisms, this
change in quantity may be reflected in a change in the
information in the genome. For higher organisms, it's correlated
with a change in information in the brain.
This understanding of transformation is
consistent with our earlier definition, as a process by which
holons associate into higher-stage holons on the same level of
existence. As we saw in Chapter 3, the primary property of a
fundamental holon that enables it to associate into higher
stages is the ability to communicate with other holons. Thus the
key role of carbon atoms in all higher stages and levels of life
derives from their ability to bond with four other atoms
simultaneously. Likewise, eukaryotic cells have the ability to
form associations with other cells of their kind; and human
beings have by far the most sophisticated forms of communication
of all organisms.
The ability of fundamental holons to
communicate with other holons of their kind, in turn, is
correlated with the amount of information they have. Eukaryotic
cells have larger genomes than prokaryotes have, and human
beings and other higher vertebrates have larger brains than
those of lower organisms. The greater the amount of information
the fundamental holon has, the greater the variety of ways in
which it can communicate with other holons of its kind. So
transformation necessarily implies an increase in information.
In Chapter 3 I also discussed the important
distinction between the deep structure and surface structure of
information. To refresh our memories, deep structure is the
total informational content present in a fundamental holon. The
deep structure of the genome is the total genetic information it
contains, represented by all the genes. The deep structure of
the brain is the total amount of biological information it
contains, represented by its hard-wired anatomy, and all the
potential activities that anatomy can manifest. The deep
structure of the genome is the same for every cell of a given
species (species of unicellular or multicellular organism), and
the deep structure of the brain is likewise the same for every
organism of a given species.
The surface structure of information, on the
other hand, is the way that information is expressed in
different individual cells or different individual organisms of
the same species, or in the same cell or organism at different
times. The surface structure of the genome is represented by
those genes which are translated in a particular cell at a
particular time. The surface structure of the brain is that
pattern of neural activity occurring in a particular organism at
a particular time.
During transformative processes, changes
occur in both the deep and the surface structures of
information. However, the involvement of each type of change
depends on where in a particular level of existence evolution is
occurring. Generally speaking, the evolution of the lower stages
on any level is driven by changes in the deep structure of
information, while evolution of the higher stages is driven by
changes in the surface structure.
For example, evolution of the lowest
invertebrate organisms was associated with the emergence of
eukaryotic cells with their large genomes. These cells were very
different from the prokaryotes that first evolved on earth, and
the large genomes continued to evolve as higher invertebrates
emerged. With the evolution of the higher vertebrates, and
particularly the higher primates and our own species, changes in
surface structure became much more important. Indeed, as I
pointed out in the previous section, the total genetic
information in the cells of all the higher vertebrates is pretty
much the same. In other words, these organisms all contain very
similar genetic deep structures. Genetically, their differences
can be described only in terms of changes of surface structure,
how this information is expressed.
The same relationship can be seen on the next
level of existence, the mental. The earliest forms of social
organization among organisms are observed among the lower
vertebrates, and some invertebrate species. These animal
societies are associated with evolution of the deep structure of
the brain. The highest social forms of organization emerge only
with our species, and at this point, virtually all of the
evolution is of surface structure in the brain. Thus the deep
structure of the human brain completely evolved fifty to one
hundred thousand years ago, yet the most profound changes in
human social organization have occurred since that time.
Another important lesson we learned earlier
is that evolution of social stages generally occurs on two
different levels simultaneously. While the surface structure of
the genome was perfecting itself within eukaryotic cells, these
cells themselves were evolving into more complex tissues and
organs, particularly the brain. While the surface structure of
the human brain has been evolving over the past several thousand
years, individual human beings have been forming ever more
complex societies. Note how the relationship works. On the lower
level, a change in surface structure (expression of the genome;
human language) is correlated with a change in deep structure on
the next level (the hard-wired brain; human societies). Thus the
higher stages of one level are completed while the lower stages
of the next level are formed. When the former process is
complete, the higher stages of the next level evolve,
accompanied by the lower stages of a still higher level.
To summarize, transformation, in contrast to
diversification, involves a change in the amount of
information--in the genome or the brain, depending on the level
of existence evolving. On the lower stages of any level of
existence, this change can be understood as a change in the deep
structure of the informational holon. Evolution of the higher
stages of any level of existence, in contrast, are associated
with changes in the surface structure of information. Changes in
surface structure do not involve changes in the total amount of
information, but rather in the way in which that information is
expressed.
Now, it should be apparent, we can see where
in evolution information from outside the system must come in.
For to say that changes in surface structure occur is just
another way of stating that there is a disconnect between the
amount of information at one level of existence, and that on
another. Thus the higher vertebrates evolved even while the deep
structure of their genetic information did not greatly change.
Likewise, the higher stages of human societies have evolved, and
are continuing to do so, even while the deep structure of the
human brain has not changed. So extra-genetic, or extra-nervous,
information is acquired by the evolving level in its higher
stages.
In Chapter 2 I discussed how the stages of
any level of existence could be understood in terms of
dimensions of both space and time, with the temporal dimensions
becoming more apparent in the higher stages. Speculatively,
then, we can say that the dimensions of space are created by the
informational holon within the evolving level (genome or brain),
while the dimensions of time are created by information outside
of this holon. In the cell, the temporal dimensions are
those above the three-dimensional polymer, that is, in
macromolecular structures that are composed of many proteins or
nucleic acids. These holons are far better developed in
eukaryotes than in prokaryotes. In the organism, the temporal
dimensions are found in the more organized nervous systems,
generally associated with vertebrates.
In both types of holons, perception of time
first becomes apparent. Thus eukaryotic cells, as I discussed in
Chapter 3, exhibit adaptive responses that allow themselves to
adjust to a prolonged chemical stimulus. Vertebrate organisms,
likewise, have a much more developed sense of time than
invertebrates, being able to communicate by means of behavior
patterns.
In conclusion, all transforming processes
involve the acquisition of information, so that the higher stage
has a greater quantity of information than the lower stage. In
the evolution of the first three stages of a new level, more or
less, this information is accumulated in an informational holon,
such as the genome or the brain, and is manifested as new
spatial dimensions. In the higher, concluding stages, this
information is outside the system, and is manifested as new
temporal dimensions.
Transcendence
Transcendence is the third and final phase of
evolution, completing a new level of existence. To appreciate
how this process differs from transformation, we can review the
ways in which a new fundamental holon differs from its component
social holons:
1) It is autonomous, capable of existing
outside of higher-order holons.
2) It can reproduce itself.
3) It preserves all the properties of the
social holons within it.
The first two of these features are obviously
very closely related. If a holon is to be capable of existing
autonomously, it must be able to reproduce itself. Leaving aside
point 3) for now, I will therefore consider the emergence of
reproduction as a central feature of transcendence. In fact, I
will use it to offer a new definition of transcendence: the
process by which a group of holons acquires the ability to
reproduce itself.
We also saw in part 1 that reproduction
requires information. Since a fundamental holon contains and
organizes all the stages on the level of existence below it, all
these holons must be reproduced in the process of reproducing
the fundamental holon. This can only be done if the latter has
information about them, information that, in effect, enables it
to duplicate each of them. So transcendence, like
transformation, involves evolution of information.
In the case of transcendence, however, an
entirely new informational holon must be created. The
transcendent process associated with the emergence of cells
required evolution of the genome. The emergence of organisms
required evolution of the brain. Emergence of a planetary
super-organism or super-culture, I have argued, requires a new
informational holon for that level. So while diversification is
associated with a change in the quality of information, and
transformation with a change in the quantity of information,
transcendence is a change in the organization of information.
Evolutionary biologists have long recognized
the problem this poses, particularly in the evolution of the
cell. Classically, the problem has been posed in terms of a
choice, between function and reproduction. Function includes all
the properties that allow a holon to grow and maintain itself,
properties that I have defined as assimilation, adaptation, and
communication. In the cell, all these properties are ultimately
controlled by enzymes, protein molecules that catalyze all the
metabolic reactions. Reproduction, on the other hand, is
controlled by DNA, which contains the information needed to make
these enzymes, and which is capable of reproducing itself.
It's not too difficult (relatively speaking)
to conceive of scenarios in which the functional or reproductive
aspects of cells evolved separately. In a later chapter, we will
see how enzyme molecules can form self-sustaining networks in
which the products of one reaction catalyze another reaction. If
a group of such enzymes were to become enclosed by a lipid
(fatty) membrane, isolating them from the surrounding
environment, they would constitute a very primitive cell, in a
functional sense.
We can also imagine the early evolution of
nucleic acids, formed by polymerization of nucleotide bases. It
has been suggested that these bases may have been adsorbed to
the surface of silica clays, providing a way of bringing them in
close proximity for the polymerization process to occur (Rao et
al. 1980; Friebele et al. 1980; Coyne 1985). Once these
primitive nucleic acids were formed, they could then provide the
surface, or template, for their own reproduction, just as DNA
does in cells today.
Neither of these two scenarios is really as
simple or as unproblematic as I have implied. I have glossed
over substantial difficulties in working out the evolutionary
details of these processes. The really difficult problem,
however, is attempting to understand how these two kinds of
evolutionary events, functional and reproductive, could have
been brought together. A primitive sac of enzymes can't
reproduce itself; a primitive nucleic acid molecule can't code
for enzymes. Somehow, the information represented in the
functional cell had to become encoded in the nucleic acids. Only
when this occurred could we speak of the emergence of genuine
cells.
In Chapter 3, we saw that the computer
scientist John van Neumann posed this problem in terms of
transcription and translation. When nucleic acid molecules
reproduce themselves, a process of transcription occurs.
Whatever information the nucleic acids contain is copied, to
form a duplicate nucleic acid molecules. When this information
is used to synthesize a protein, on the other hand, a process of
translation occurs. Thus the cell must not only encode all its
information into nucleic acids, but it must be able to read this
information in two different ways. In computer science terms, it
must be able both to copy the program and to run the program.
In the past decade, a major discovery has
been made that may shed some important light on this problem. A
class of nucleic acid molecules have been identified, call
ribozymes, that have the ability to act as enzymes; in other
words, they possess both functional as well as reproductive
properties (Zaug et al. 1986). In the past few years, many
different kinds of ribozymes have been discovered. Furthermore,
scientists can actually create new ones in the test tube by in
vitro evolution. In this process, ribozymes are allowed to
reproduce, with the new generations being subjected to
artificial selection processes (Joyce, 1992). Such studies
suggest that ribozymes might once have existed possessing a wide
range of catalytic properties.
Many scientists now believe that such
functional RNA molecules were a key transitional stage in the
evolution of cells. Several hypothetical scenarios have been
proposed by which cells might have emerged from such an "RNA
world" (Horgan, 1996; deDuve, 1996; Schwartz, 1998; James and
Ellington, 1998). These scenarios, like others for early
evolutionary events, are largely untestable, and not all
scientists accept them. It has been pointed out that a few
proteins can reproduce, such as the recently-discovered prions
that are thought to transmit mad cow disease and several other
major disorders (Prusiner 1997, 1998). Such proteins conceivably
might have performed as the missing link between purely
functional assemblies of enzymes and reproducing cells.
Nevertheless, the preponderance of evidence
seems to favor a role for ribozymes. While it's true that the
genetic material of all modern cells is DNA, rather than RNA,
the latter plays an essential role in translation, that is, the
synthesis of proteins. Indeed, three different kinds of RNA are
involved in this process: messenger RNA, which transcribes, or
copies, the DNA sequence; transfer RNA, which mediates the
interaction of specific amino acids with specific three base
sequences in the messenger RNA; and ribosomal RNA, which acts as
a sort of scaffolding, or structure site, for the process of
amino acid polymerization to take place. Furthermore, RNA
molecules can regulate DNA transcription, and many
ribonucleotides play key roles as cofactors in enzymatic
reactions (Strickberger, 1996). So RNA in many respects appears
like a "vestigial organ" of the cell, a once highly functional
molecule that now plays a less central role.
In any case, while it will be very difficult
to establish just how early cells evolved, the concept of a
single type of molecule embodying the two key aspects of
function and reproduction is a very powerful one. The problem of
unifying or synthesizing these two kinds of processes is not
restricted to cells. It's a key problem in transcendence at
other levels of the holarchy as well. To the extent that
processes on different levels are analogous to one another, we
might expect that this problem would be solved in a similar
manner at other levels.
Consider the process of transcendence on the
next level of existence, that is, the emergence of the first
multicellular organisms. We have seen that the brain is the
biological level analog of the genome on the physical level.
Just as the genome controls reproduction in the cell, the brain
controls reproduction in the organism. It obviously does not do
this in a physical sense--the reproduction of tissues and organs
in a new organism--for this, too, is controlled by the genome.
The brain controls reproduction in a biological sense--that is,
it contains the information needed to regulate the functions of
these tissues and organs--to control respiration, heartbeat,
gastrointestinal processes, muscular movements, and so forth. In
all but the most primitive organisms, a brain or centralized
nervous system controls the activity of all the body's organs by
sending regular messages to them. Thus, I suggested before, all
of the latter--the major organs and tissues (more precisely the
activity of these organs and tissues)--are the biological analog
of enzymes.
We can imagine that during the evolution of
the first organisms, there were probably both multicellular
functional arrays--primitive organisms capable of digesting
food, circulating nutrients, movement, and so forth--as well as
multicellular reproductive arrays--a primitive nervous system
capable of encoding information in the patterns of connections
among excitable cells. As with evolution of the cell, one could
imagine either type of array evolving independently. That is,
some cells could have associated into primitive organ-like
structures capable of carrying out certain functions, and other
cells could have evolved excitable properties allowing them to
associate into primitive nervous systems. The key issue is how
both types of holons evolved together, in such a way that the
nervous system contained patterns of information capable of
regulating the activity of the organs. This problem, I suggest,
was just as difficult to solve as the genome versus protein
problem, and I further suggest that it was solved in the same
way as the emerging cell may have: by the emergence of a hybrid
holon capable of both function and reproduction.
What would such a hybrid holon be like? As a
ribozyme can both contain reproducible information (nucleotide
base sequences) and function (catalytic enzyme activity), a
hybrid biological holon would be capable of functioning
simultaneously like a nervous system and a functional organ.
Such a holon still exists in most organisms today, including
ourselves: the heart. Though the activity of the heart, like
that of other internal organs, is regulated by centers in the
brain, it also has some ability to regulate itself. Much of this
activity is controlled by the sinus node in the atria, muscle
tissue which has excitable properties like nervous tissue. In
addition, in the absence of control from the sinus node, other
parts of the heart can also control its activity. This is why
the heart can be completely removed from the organism, severed
from all of its connections with other organs, and still beat.
In addition to the heart, there are other
internal organs with the ability to function to some extent in
the absence of control from the central nervous system. The
intestine is innervated by a very primitive kind of nervous
system that induces rhythmic contractions in the smooth muscle
lining this organ (Gershon, 1998). The vas deferens, part of the
male reproductive system, is controlled by a somewhat autonomous
network of nervous connections. I suggest that in such tissues
and organs we can see vestiges of a critical stage in the
evolution of organisms, when primitive organs emerged which were
self-regulating, capable of both function and the ability to
control this function through reproducible patterns of activity
in excitable cells. This stage was transcended only when a more
centralized and more specialized nervous system emerged that
gradually superseded the local form of control.
What about our own level of existence, the
mental level? The scheme of transcendence I am developing here
predicts that there should be, or later will be, a similar
distinction between holons storing information in a reproducible
form and holons exhibiting functional activity. Given that our
level of existence has not completed its evolution, we would not
expect these holons to be fully emerged. But surely they are
beginning to emerge, on the one hand, in the cumulative
knowledge of modern societies--in oral, written and printed
language--and on the other, by technology, in all its forms.
Language, like informational holons on lower levels of
existence, is a medium which can not only reproduce itself, but
in principle can contain the instructions for reproducing--not
in a physical or biological sense, but in a mental
sense--all of society. Technology, on the other hand, like
functional holons on lower levels, has the potential to
translate or execute these instructions.
It also appears that the hybrid holon
necessary to complete the transition to a new level of
existence--a holon that combines knowledge with technology--is
also beginning to emerge. This is the computer culture,
including not only computer hardware and software, but the human
beings who use them. This culture has the potential, for the
first time in human history--i.e., in the evolution of our
level--to combine reproducible information and technological
production in a single type of organization. Thus computers can
not only store all the information needed to reproduce society,
but can increasingly function as the technology that executes
these instructions as well. We see this in the fact that more
and more of our technology today incorporates computers directly
into its design.
Evolution and Information
To summarize and conclude this chapter, we
have seen that evolution throughout the holarchy can be
considered to take place in three different phases:
diversification or translation; transformation; and
transcendence. Diversification is the process by which different
kinds of holons on the same stage of existence are created.
Transformation is the process by which higher stages of
existence are formed from lower stages, on the same level.
Transcendence is the process by which a new level of the
holarchy emerges.
The ultimate goal of evolutionary theory, it
seems to me, should be to develop a framework in which all three
kinds of processes can be understood. Part of this framework
lies in the recognition that processses on one level of
existence are analogous to those on another level. I have
already presented abundant examples of such analogies in Part 1
of this book. In succeeding chapters in this part, we will
examine the extent to which these analogies are also evident in
evolutionary processes.
Further unification of evolutionary theory,
however, will require a concept with which we can understand all
types of processes. I believe this concept is most likely to be
information. I have tried to show in this chapter how each type
of process--diversification, transformation, transcendence--can
be understood in terms of information. Thus diversification is a
process by which the quality of information changes;
transformation is a process in which the quantity of
information changes; and transcendence is a process in which the
organization of information changes.
To be sure, for information to play a truly
central role in evolutionary theory we will need both to
understand it better and to measure it better. As we learn more
about how DNA sequences in the genome and neuronal connections
in the brain are translated into the phenotypes of the cell and
organism, the nature of information in living things should
become clearer to us. At the same time, advances in information
theory may help us understand what the relevant genetic and
neural processes are that we need to examine.
Of course, it's one thing to say that
information changes in certain ways during evolution; it's
another to understand how this change occurs. We will now
consider some evolutionary theories, to see how well they are
capable of explaining the processes I have just described. We
will begin with Darwinism, still the central theory of evolution
today.
Go to Footnotes
Go to Chapter Summary