3. TRANSLATION , TRANSCRIPTION AND COMPRESSION

"We have processes inside one E. coli that do what we used to think required many complete nerve cells. E. coli is about the size of a single spine of a pyramidal cell [a type of neuron], and each pyramidal cell has about five thousand spines; its total volume is equivalent to about a quarter of a million E. coli, so these facts should bring about a revolution in the way we think about computation in the brain."

-Horace Barlow1

 

"I have a hunch that there's some deep way in which IBM and E. coli know their worlds in the same way."

-Stuart Kauffman2

 

In Chapter 2, we saw that the forms of existence within a cell--beginning with atoms and comprising ever more complex molecular structures--can be said to constitute a single level of existence, which is usually called the physical level, or the level of matter. In this chapter, we will examine the next level of existence, the biological, beginning with the cell and culminating with the organism. This level is also called the level of life.

 

Viruses: Agents of Evolutionary Change?

Cells are generally considered the simplest form of existence that can genuinely be said to be alive. Before discussing them, however, I want to say something about viruses, a still simpler form of existence composed of just nucleic acids (DNA or RNA) and a few proteins. Viruses are usually considered an intermediate form of life, more complex than any large molecule, yet lacking most of the properties of cells. Most scientists believe, however, that viruses first appeared after the evolution of cells, and so are not a true transitional form between the latter and complex molecules.

More specifically, the predominant scientific view is that viruses evolved from intracellular pieces of DNA that are to some degree separate and autonomous from the main genetic material (Campbell 1981). These semi-autonomous sequences of DNA include plasmids, found in many bacteria and in simple eukaryotic cells like yeast; and transposons, found in the cells of some multicellular forms of life, including our own species. Plasmids are, in effect, smaller, satellite sequences of DNA, which can be transferred from one bacterial cell to another. Transposons are sequences within the genome that can under some circumstances change their position, moving or "jumping" from one area to another of the genome. Thus both plasmids and transposons are mobile genetic elements, travelling pieces of information that can modify the DNA sequences within cells.

Viruses in some respects combine the features of both plasmids and transposons. Like the former, they can be transferred from cell to cell; like the latter, they can integrate themselves into the genome of the host cell. In this way they spread from cell to cell and organism to organism, using the genetic material of the host cell to reproduce themselves. One could speculate, therefore, that they originally evolved as a means of transferring information from one cell to another.

Lynn Margulis is best known for her theory, now widely accepted, that mitochondria, subcellular organelles found in all eukaryotic cells, evolved when these cells assimilated smaller prokaryotes like bacteria (Margulis 1971). More recently, she has suggested that plasmid transfer among prokaryotes makes possible a virtually world-wide web of genetic information, shareable among all bacteria on the planet (Margulis and Sagan 1986). Viruses might be regarded as part of a similar web among eukaryotic organisms. Though viruses are now considered, at best, harmless, and at worst, of course, lethal to organisms, at one time they may have contributed to the evolution of the latter by bringing new information into the genome. That is, they would pick up a piece of DNA from one organism and transfer it to another.

When multicellular organisms were first evolving, this kind of process might have been capable of creating a great deal of genetic diversity. However, it's not at all clear how much evolutionary change would be possible from such a mechanism today. Most organisms reproduce sexually, using gametes that are segregated from the other cells of the body. Thus viral information transfer into the somatic (non-sexual) cells of organisms would have no effect on subsequent generations of organisms. Furthermore, viruses today are usually species-specific. A given type of virus generally does not infect more than one kind of organism, or several closely-related organisms (Singer and Berg 1997), though given their great capacity for change through mutation, this is not out of the question.

Viruses might affect the evolution of our own species in another way, however. Recent research has revealed that susceptibility to AIDS, one of the most devastating viral diseases to have afflicted human beings, is correlated with the presence of certain receptor proteins on the surface of immune cells. That is, individuals with certain types of these receptors are less likely to become infected with the virus than individuals with a different kind of receptor (Michael et al. 1997). It's not inconceivable, therefore, that AIDS, or some other viral disease, could exert a powerful selective force for human beings of a certain genetic composition. To be sure, for this to occur, a much larger fraction of people would have to die than the already enormous numbers that have so far. In addition, for the selection to have much significance biologically, the genetic composition of the survivors would have to confer on them differences in properties other than simply resistance to a particular virus. AIDS, therefore, probably is not nor will it become a selective agent in this sense. However, as the population density of our species increases, together with a still greater increase in interactions among people all over the planet, the emergence of new, still deadlier viral diseases is a serious possibility (Preston 1995; Garrett 1995).

Furthermore, many viral diseases, including AIDS, could result in selection at levels of existence other than the biological. By significantly reducing the number of people in particular areas of the world, these scourges may drastically alter the nature of the societies these people live in. Indeed, in some places, such as parts of Africa, this has already occurred, where a large fraction of the most productive members of society have died (Bennett and Blaikie 1992; Bond et al. 1999). Even developed countries like the U.S., where a much smaller fraction of the population has been affected, have experienced significant social change, for example, in new attitudes towards sex (Feinleib and Michael 1998). These observations suggest that widespread viral diseases might have a major impact on the evolution of the higher, social stages of the holarchy. I will discuss human societies in the following chapter, and in Part 2 of this book, we will examine how evolution may occur at other levels of existence besides the genetic/biological.

In conclusion, from the holarchical point of view, some viruses might be considered to be a component of organisms, or of societies of organisms. In the early stages of evolution, they may have transferred genetic information from one organism to another, greatly increasing the diversity of life. In our present era, their impact on evolution is most likely greatest in the social structure of our societies. Nevertheless, they are not an autonomous form of life like cells. The latter, which I will now discuss, are much more fundamental to the development of holarchy.

 

Properties of Cells

We can see in cells many of the same properties that we see in ourselves and other living organisms, albeit in a very rudimentary form. As I suggested in Chapter 2, most of these properties can be grouped into one of three fundamental features: assimilation or growth; adaptation or self-maintenance; and communication. We will briefly consider each of these properties as they are manifested by cells, then look at a fourth property, new in the holarchy with cells: reproduction.

1. Assimilation (growth). All cells have the ability to grow, that is, to assimilate substances or energy from their environment. This quality distinguishes them fairly clearly from lower-order holons. As we saw in Chapter 2, atoms and molecules have the ability to grow in a very simple sense, by assimilating other atoms or electrons. However, this growth is limited. In the case of atoms, one or a few electrons are incorporated, at which point no further assimilation can occur. Molecules, especially large, complex ones, may incorporate many atoms, but in a sense, this is not true assimilation, because the identity of the original molecule is not retained. That is, in the process of incorporating atoms, a molecule becomes some other kind of molecule.

In contrast, a cell incorporates many kinds of substances without changing its basic identity as a cell. These substances include most of the basic kinds of holons found within cells. That is, cells can assimilate atoms, such as sodium ions; small molecules such as amino acids and simple sugars; and in some cases, macromolecules such as proteins, and even small organelles. Some unicellular organisms, such as Paramecia, can even ingest other cells, and as I noted earlier, certain organelles within cells seem to have originated from such an assimilation process. In all cases, the substances are transformed--or digested as we would say of the analogous process within organisms--within the cell so that they become part of the cell.

This process of transformation clearly distinguishes assimilation in cells from the primitive process we identified in atoms in Chapter 2. Whereas an electron that is captured by an atom retains its basic identity as an electron, a substance ingested into a cell is usually converted into another substance. This transformation may result in either lower-order holons, through a process of metabolism or degradation, or higher-order holons, through synthesis. An example of the first process is the conversion of a sugar molecule into carbon dioxide and water, while the second is illustrated by incorporation of an amino acid molecule into a protein. Only atoms, like sodium and calcium ions, retain their original properties following assimilation into the cell.

Thus when a holon is assimilated into a cell, it may move, so to speak, in one of three possible directions in the holarchy: down to a lower form of existence; up to a higher form; or it may retain its original identity and position in the holarchy. Processes that involve a downward movement ordinarily result in the extraction or accumulation of energy. When sugar molecules are metabolized, energy is released from the chemical bonds that are broken, and some of this energy is captured and stored by the cell. When the movement is upward, in contrast, energy is required, to form the new chemical bonds that result in the synthesis of a higher order holon.

This suggests that energy is a useful way to understand the concept of directionality in the holarchy, that is, to distinguish up from down. If energy is required to create higher-order holons, and is released during the formation of lower-order holons, it follows that higher-order holons have more energy than lower-order holons. I will discuss this important idea further later in this chapter.

2. Adaptation (self-maintenance). In Chapter 2, I defined adaptation as the interaction of a lower-order holon with a higher-order holon. A cell in an organism, for example, adapts to conditions surrounding it, which are part of the organism, a higher form of existence. When considering autonomous cells, however, the higher-order holon may need to be understood in fairly broad terms. A single-celled organism swimming in a body of water adapts to this surrounding liquid environment, which may not actually be a higher-order system. However, in nature, the environment of any cell, in the broadest sense, is part of the earth, which as we will see later, may be viewed as an emerging higher-order system.

A very common example of adaptation found in many cells occurs in response to a change in sodium ion (salt) concentration. All cells have a characteristic concentration of sodium inside them, which is vital to their growth and functioning. Many cells, though, inhabit an environment in which the sodium ion concentration is much higher than that within the cell. In a simple physical system, the sodium outside the cell would diffuse into the cell, until the concentration of this ion inside and outside of the cell was the same. Cells, however, have special molecules called "sodium pumps" on their surface which actively extrude sodium ions from the cell, and so maintain their low internal concentration of this ion (Matthews 1998). This process requires energy.

Another example of an adaptive process in cells is provided by their response to chemical substances used in cell-cell communication. Many cells in the body respond to special chemical agents, such as neurotransmitters or hormones, which induce the cell to grow, to reproduce itself, or to secrete its own chemical messengers. Sometimes, however, the cell may be overexposed to the substance. Under these conditions, the cell may alter its sensitivity to the substance, so that it does not become overwhelmed by it. It does this by reducing the number of receptor molecules on its surface that interact with the chemical messenger, or by altering the degree of fit or affinity that the receptor has for the messenger (Law et al. 1984). This process is called desensitization.

Overexposure of a cell to a chemical messenger may result from either a higher than normal concentration, or to a normal concentration that is maintained for a longer than normal period of time. The latter situation is particularly interesting. Under these conditions the cell is responding to not its immediate environment--the presence of a certain substance--but to the past history of that environment. Its behavior is shaped by experience.

We can call this kind of adaptation a primitive form of learning. What distinguishes this kind of adaptation from the simpler forms a cell is capable of is the dimension of time. Whereas in the simpler form of adaptation a cell responds only to the presence of a substance in space, in this form it's responding to this presence over a particular length of time. In other words, the cell is capable of recognizing a distinction not only in the physical dimension of some holon, but also its temporal dimension. As I discussed in the previous chapter, the ability to function in temporal dimensions is usually a property of higher stages of existence. Cells that can learn in this sense are usually found in organisms, and particularly in the brain, where they participate in these higher stages of multicellular organization. I will return to this point later.

3. Communication. Communication, we saw in Chapter 2, is the property of holons manifested in their interaction with other holons of the same or similar type. Thus atoms communicate with each other by forming chemical bonds between themselves. Communication, defined in this way, is critical to the development of the holarchy, for it allows higher-order holons to be created from simpler ones. Atoms communicating create molecules; molecules communicating create polymers; polymers communicating create larger molecular structures.

Communication at the level of the cell is, again, more complex, more sophisticated, than communication at the level of atoms and molecules. There are a variety of ways cells can communicate with each other. The simplest forms of communication involve direct physical interaction between the cells, as occurs in all tissues in the body. Even this interaction is not really simple, however, as it involves some very specialized macromolecules on the surface of each cell, which recognize and bind to each other (Edelman 1984; Lander 1989; Tessier-Lavigne 1996). This binding often triggers internal metabolic processes within the interacting cells. Cells may also communicate physically by means of gap junctions, tiny channels that connect adjacent cells (Dulbecco 1987).

As I noted in the previous section, some cells can also communicate at a distance, by secreting certain chemical substances or messengers that trigger specific responses in other cells. This kind of communication may occur over very short distances, as occurs between neurons during synaptic transmission, and over very great distances, such as when cells release hormones into the blood stream that act at cells in other parts of the body. Sometimes this communication is hierarchical, as when cells in the hypothalamus in the brain release hormones that activate cells in the anterior pituitary, another brain nucleus. The pituitary in turn, releases another set of hormones that activates tissues such as the adrenal glands, the sexual organs, and the thyroid (Dulbecco 1987). Hierarchical communication also is very commonly carried out by nervous transmission. For example, cells in the brain send signals to cells in the spinal cord, which in turn send signals to muscles.

As I discussed in the previous chapter, hierarchical organization of this is non-nested, occurring between holons of the same type which participate to different extents in higher, social holons. As with nested or holarchical relationships, however, this organization is not completely unilateral. While the brain sends signals to the muscles via the spinal cord, we also know that the activity of the muscles influences the brain, through feedback loops (Baker 1999).

In Chapter 2, we saw that there was a fundamental division between reactive atoms, which are capable of forming bonds with each other, and inert atoms, which do not. Only reactive atoms can create molecules and thus become part of higher stages and levels of existence. The same fundamental division exists on the biological level, between prokaryotes and eukaryotes. Prokaryotes, which include most bacteria3, live a fairly autonomous existence. While some prokaryotes are capable of interacting with one another, through chemical signalling, and may even form loose aaggregations of cells, only eukaryotic cells are capable of the highly complex, hetarchical associations that make up organisms. Thus eukaryotes, like reactive atoms, are highly communicative.

What is it about eukaryotes that makes them so adept at communicating, at interacting with other holons of the same type? One of the major differences between prokaryotic cells and eukaryotic cells is that the latter have much larger and more complex genomes. The genome is the cell's repository of information; this is contained in the sequences of DNA, which code for all the proteins in the cell. As we just saw, when one cell communicates with another, it generally does so by presenting a certain kind of molecule to the latter, either directly in physical association, or indirectly, as through communication by hormones or neurotransmitters. The second cell must recognize this molecule with another molecule of its own. A large genome thus enhances the ability of cells to communicate with one another, by increasing their repertoire of signalling molecules. Every molecule that a cell uses to communicate with another cell must be encoded by a distinct gene4, so the more genes a cell has, the greater the potential variety of communicative interactions it can make with other cells. Indeed, it appears that a great deal of the extra genetic information present in eukaryotes is devoted to communication. "Half the genes that we have are involved in intracellular communication,"5, estimates molecular biologist Tony Hunter--intracellular communication being the final step in intercelluar communication.

Not only is the genome of the eukaryotic cell larger than the genome of prokaryotic cells, but there appears to be a further division among eukaryotic cells. Bernardi (1993) has pointed out that the genomes of the highest vertebrates, including birds and mammals, are much larger than the genomes of lower vertebrates and invertebrates. He calls the latter the paleo- (evolutionarily old) genome and the former the neogenome, a distinction that suggests a parallel at the next level of existence between the paleocortex and neocortex in vertebrates. In the next chapter, we will see that the brain in fact plays an analogous role in the organism to that the gene plays in the cell. For now, we just observe that there is a correlation between the size of the cell's genome and the ability of the cell to form complex multicellular holons.

I also pointed out in Chapter 2 that among reactive atoms, carbon is of primary importance, because of its ability to form chemical bonds with four other atoms simultaneously. This property is what makes carbon the so-called building block of life. The analogous holon on the biological level is represented by the neuron, or nerve cell. Neurons are the most communicative of all eukaryotic cells, because, like carbon atoms, they can interact simultaneously with many other cells.6

To summarize, there is a correlation between the amount of information a cell has, and its ability to communicate with other cells. Not only do eukaryotes contain more genetic information than prokaryotes, but eukaryotic cells of higher organisms--which have more complex multicellular organizations, particularly in the brain--generally contain more information than the genomes of cells in simpler organisms. There are some very significant exceptions to this rule, which we will consider later, but for now I want to emphasize the general correlation.

As we will see in the next chapter, a similar correlation holds with organisms, on the next level of existence. On that level, there is a strong correlation between the size of the brain and the extent to which the organism forms social organizations. A traditional definition of communication, of course, is the transmission of information, so this correlation should not be surprising. Some systems theorists, such as Valera and Maturana, prefer to define communication in a way that does not involve reference to information, such as "coordination of behavior."7 While this definition is quite consistent with the one I'm using here, I believe the relationship of information to communication is critical, as I will discuss at some length later in this chapter.

4. Reproduction. In addition to the trio of assimilation, adaptation and communication, cells also have the ability, of course, to reproduce themselves. This property perhaps most clearly distinguishes cells from all lower forms of existence. Atoms, molecules and complex molecular structures can't reproduce themselves. DNA, to be sure, can reproduce itself in the test tube, but this requires special conditions, including the presence of certain enzymes, which are not found in nature (Kornberg and Baker 1992).

As I pointed out earlier, each of the three universal properties of holons can be defined in terms of interaction with other holons. In assimilation, a holon interacts with a lower-order holon; in adaptation, it interacts with a higher-order holon; and in communication, it interacts with a holon on its own stage and level of existence. Reproduction (of the cell) is unique in that it's the only property of a holon that does not involve interaction with another holon. Alternatively, it's the only property of a holon that involves simultaneous interaction with holons above, below, and on the same plane of existence as itself.

What do I mean by this paradoxical statement? Consider the first definition. When a cell divides, to be sure, numerous interactions occur among the holons it contains (Campbell et al. 1999). The chromatin material in the nucleus condenses into chromosomes, which pair up along the mitotic spindles, molecular structures that hold the chromosomes in place. The nucleus then divides, along with the rest of the cell. The molecular details of these processes are quite complex. But from the point of view of the reproducing entity, the cell, these interactions are with itself. In this sense, its reproduction does not involve interaction with any "other" holons.

On the other hand, cells in the organism (and even cells living autonomously outside of organisms) normally don't divide in an uncontrolled fashion--when they do, it's generally a sign of a disorder, for example, the growth of a tumor. Indeed, for any eukaryotic cell to divide, it must undergo a characteristic sequence of processes called the cell cycle. This cycle includes meiois, or halving of the chromosome number, as well as cell division into two daughter cells. At several places in this cycle, called checkpoints, the process can be aborted, if certain conditions aren't fulfilled (Murray and Hunt 1993; Stein et al. 1999).

What sort of conditions must be fulfilled? The division of cells within healthy organisms is often regulated by the tissue they compose. For example, certain growth factors (hormones or related molecules) secreted from surrounding cells may signal the cell to divide; if these factors aren't present, it won't divide (Stein et al. 1999). Cell division is also regulated by direct, physical contact with neighboring cells. This is important in preventing a tissue from growing too large. So reproduction of the cell involves both adaptation (interaction with a higher-order holon, which as we will see shortly, is represented by tissues and organs), and communication, interaction with other cells. Furthermore, reproduction of cells is also generally triggered by size--the cell must double in its mass in order to produce two daughter cells of its orginal. So reproduction is also a process of assimilation, a means of allowing it to continue incorporating lower holons into itself. As Teilhard de Chardin, poor on explanations but sharp on what needs explaining, put it: "the cell, continually in the toils of assimilation, must split in two to continue to exist."8

So in an important sense, reproduction of a cell involves all three fundamental processes that holons engage in--assimilation, adaptation and communication. In Chapter 2, however, we saw that any interaction among holons can be thought of as involving all three of these processes, with the particular one defined depending on the holon's point of view. When an atom and a molecule form a bond, for example, assimilation occurs from the point of view of the molecule; adapatation occurs from the point of view of certain electrons in the atom and in the molecule; and communication occurs from the point of view of the atom.

Incorporating this notion into the preceding discussion, we could say that reproduction is a process by which the holon's point of view, or identity, is expanded, in such a way that it participates in all three of the other processes. When a cell reproduces, it becomes an entity which is capable of perceiving itself simultaneously as a) assimilating, which is the immediate trigger or cause for reproduction; b) communicating, which gives it permission to reproduce; and 3) adapting, which allows no other response. To put it another way, assimilation makes reproduction necessary; communication makes it possible; and adaptation makes it sufficient.

 

Transcription and Translation: the Deep Structure and Surface Structure of Information

So far, I have discussed reproduction of cells in very general terms, as a series of interactions with other holons. Now we will consider some of its specifics. In particular, we will examine the role of information, which is a key concept in the process.

About half a century ago, the physicist John von Neumann (with the help of Stanislaw Ulam) developed a scheme by which a computer could reproduce itself. Von Neumann's work was the beginning of a theory called cellular automata (Casti, 1992; Wolfram, 1994), which I will be discussing in more detail in Part 2 of this book. What is relevant to our discussion here is how van Neumann solved this problem. At the outset, he recognized that since a computer can, in principle, carry out any procedure that can be broken down into a series of logical steps, it was not difficult to write a program which, when run on a computer, would enable that computer to construct a copy of itself. The program would tell the computer how to make each component (presumably using robots or other forms of automated technology), and then how to assemble these components into an identical computer.

That part is straightforward enough. The more difficult problem is that if the duplicate computer is to be able to copy itself, too, it would have to have this program as well. That is, the program that the first computer ran in order to construct the second computer would somehow have to go into the latter's construction. The second computer, in addition to being a complete hardware copy of the first, would also have to have access to the program that the latter ran in order to duplicate itself. This would allow the second computer, in turn, to reproduce itself, by operation on the program in the same way that the first computer did.

This problem led von Neumann to grasp a key principle about reproduction using programmed instructions. The program that enables a computer to reproduce itself must be run, or operated upon, in two different ways. First, it must be translated, that is, its rules followed to create the duplicate computer. And second, it must be transcribed, that is, copied, so that the duplicate computer also has this program. Conventional computers only translate programs; they generally don't transcribe them. So van Neumann's computer had to have a special construction so that it could distinguish between the two processes, and know how, and when, to change its mode of operation from one to the other.

The insightfulness of von Neumann's reasoning became clear just a few years later, when biochemists discovered that cells and organisms reproduce themselves in a manner that follows the same basic principles. The DNA in the genome of every cell is the program that allows the cell to reproduce itself, and it, too, is operated on in two different ways. This program is translated by the synthesis of messenger RNA, which in turn instructs amino acids to join in a particular sequence to form all the different proteins of the cell. This is what enables a cell to create a copy of itself. The DNA is also transcribed, by the pairing of nucleotide bases that allows one sequence of DNA to make an exact copy of itself. This provides the duplicate cell with the ability to reproduce further.

The reproduction of cells and organisms, however, differ somewhat, both from each other and from von Neumann's reproducing computer, with respect to the way these two processes are emphasized. When a cell reproduces itself, it divides in half (mitosis), so that its contents are distributed equally between the two daughter cells. This process is therefore primarily one of transcription. The DNA in the genome makes a copy of itself, with one of each of the two total copies going to each daughter cell. The two daughter cells must, of course, translate this DNA to replenish their supplies of proteins, but this translation process is not an essential part of the reproduction of cells--as it would be in von Neumann's computer. In the cell, translation comes after reproduction, as a means by which the cell replaces components that die or turn over.

When an organism reproduces itself, on the other hand, it creates a single cell (the fertilized egg) that reproduces itself many times. The resulting progeny cells then differentiate into all the various tissues of the body; some cells become muscle tissue, some become liver tissue, some become brain tissue, and so on. This differentiation process occurs by the translation, or as molecular biologists say, expression, of different portions of the genome. Muscle cells express one set of genes; liver cells express another set of genes; brain cells express still another set of genes. Thus reproduction of the organism, to the extent that it's more than just reproduction of the cell, is primarily a process of translation.

Another way to understand reproduction of the organism is as a process of reproduction and communication of cells. It begins with reproduction of the cell (transcription), and is followed by communication of the dividing cells with each other. This allows us to see that communication of cells is primarily a process of translation of information.

So while both reproduction and communication of the cell involve operation on the genome, they operate on it in different ways. Reproduction of the cell is primarily a process by which the genome is transcribed; communication is primarily a process by which the genome is translated. These two processes, in effect, point the genome in two directions, towards the stages below it, and to the level above it. On the one hand, the genome contains all the information necessary to organize or actualize all the stages below it, through the synthesis of all the proteins in the cell. This is the information transcribed during reproduction of the cell. On the other hand, the genome also contains the potential to create a still higher fundamental system on the next level of existence. This is the information translated during communication of cells.

The information in the genome which the cell makes use of in translation and transcription corresponds, respectively, to its deep structure and surface structure. Every cell in the organism contains all the genetic information that every other cell contains. The sum total of all this genetic information is the genome's deep structure. When a cell reproduces, it transcribes this information; that is, it reproduces the genome's deep structure. But cells in different parts of the body differ, as I just noted, according to which genes they translate or express. Cells in the heart express certain genes, and don't express certain other genes; cells in the liver express a different set of genes; and so on. Furthermore, even a particular type of cell may express certain genes at one time, other genes at another time. The particular pattern of genes expressed by any given cell at any given time represents its genetic surface structure. When a higher, multicellular stage of existence is formed, through reproduction or evolution of an organism, the cells translate the genome, that is, they express its surface structure.

To reiterate, the deep structure of the genome is what allows the cell to reproduce itself, by a process of transcription; this process actualizes the physical stages of existence within another cell. The surface structure of the genome is what allows the cell to interact with other cells, by a process of translation; this process allows cells to communicate with one another, crreating higher-order holons containing the cell. As we will see in the next chapter, an analogous role is played in the organism by the brain, which also has a deep structure and a surface structure. And in the second part of this book, when I discuss evolution in the holarchy, we will see that each type of structure can change--in the cell and the organism--and that depending on which does change, a different evolutionary process results.

 

The Higher Biological Stages

Keeping this in mind, let's now look at the higher stages of the biological level, the new holons created by the process of communication among cells. Just as we saw, in Chapter 2, that higher physical stages were created by atoms combining into molecules, which in turn combined into more complex molecules, we can see an analogous organization of cells. We all know that cells in the body form tissues, and tissues form organs. These terms, however--tissues and organs--though useful for a very general discussion of the kinds of holons within the organism, are rather imprecise. A closer examination of the anatomy of the human organism--which is the most highly evolved organism, and should therefore feature all the possible biological stages, in their greatest degree of development--suggests that we can again identify a series of holarchically organized holons that feature increasingly more complex associations of cells (Table 3 ).

As with the physical level, each stage consists of a homogeneous group of the holons directly below it, and each higher stage has emergent properties not found in the lower. Biologist Rudolf Raff, who refers to biological holons as modules, suggests that they have four general properties (Raff 1996): 1) a genetically discrete identity, that is, each cell in a module expresses an identical set of genes (same genetic surface structure); 2) they are composed of lower-order modules and are in turn part of higher-order modules--i.e., are true holons; 3) have a distinct physical location in the organism; and 4) can have various degrees and kinds of interconnectivity.

Let's now consider the relationships of these holons in light of what we learned in Chapter 2. Recall that atoms and cells, unlike the stages between them, are capable of an autonomous existence; that is, they can survive outside of larger-order holons. The same is true of organisms, obviously, and conversely, is not true for the stages between cells and organisms. The various kinds of multicellular holons presented in Figure 2 are found only within organisms; they can't exist on their own. So in this important respect, the biological level is analogous to the physical.

Another important analogy between organisms and cells concerns the way they're organized. We saw in Chapter 2 that cells contain all the lower physical stages, both in semi-autonomous and combined form; thus cells contain free atoms as well as atoms within molecules; free molecules as well as molecules within polymers; and so on. The same is true of organisms. Within the organism are free cells, such as red blood cells, and cells combined into tissues; simple tissues, as well as simple tissues combined into higher-order tissues; and so on for other social holons. Thus the organism transcends its biological stages in the same way that the cell transcends its physical stages. In both cases, all the properties of the lower holons are preserved (in autonomous forms of these holons), alongside with the emergence of entirely new properties. Conversely, higher-order multicellular organizations do not preserve all the properties of their individual cells. Thus the ability of cells within tissues and organs to grow and divide is regulated by the tissue or organ.

We also saw in Chapter 2 that the emergent properties of the higher physical stages can be understood in terms of new dimensions, of both space and time. A small molecule exists in one more dimension than an atom; a polymer exists in two or three dimensions; and still more complex forms of molecules may exist in one or more dimensions of time as well as space. The higher biological stages can be viewed in the same manner. A complex cell unit, for example, is a one-dimensional array of cells, while tissues and organs can be understood in terms of two or three dimensions.

The dimensions of time of higher biological organization may be understood in several ways. Most basically, time is inherent in the process of cell turnover; any biological tissue or organ is constantly undergoing a process of self-renewal, in which cells die and are replaced by new ones. This renewal process gives the stage one or more dimensions of existence in time as well as in space. In fact, the dimensions of time and space are somewhat interchangeable. A simple cell unit, for example, can be understood as an array of cells in space, or as the life of a single cell--as it reproduces itself and forms progressively more cells--over time. In Raff's words, "a cell lineage can be considered a temporally connected series of cellular modules."9

In the brain, the most complex and highly developed biological stage, most cells don't divide, though there are some exceptions (Lichtmann 1999; Oppenheim 1999). In this case, however, the temporal dimensions of holons can be understood in terms of patterns of electrical activity, which change in time as well as in space. When we think, remember, feel or express emotions, engage in certain physical activities, certain patterns of nervous activity occur in the brain, which science is now beginning to explore using procedures that follow the metabolic activity of the cells involved (Magistretti 1999). These patterns result from the synaptic connections between neurons, which enable them to communicate with each other in complex networks.

As with higher-order physical stages, the emergent properties of higher-order biological stages result directly from the new dimensions in which they exist. These emergent properties include not only the various well-known functions of tissues and organs--digestion, circulation, and so forth--but greater stability and lifetime. Thus tissues and organs have a length of existence that far exceeds that of their individual cells; while the cells die, they are replaced by new ones that sustain the tissue or organ. For the same reason, multicellular holons are stable to the removal of a few of their components.

In the brain, where most cells don't die, we can't talk about different lifetimes, but we can still observe that higher-order holons function over longer periods of time. For example, groups of highly interconnected cells in the cerebral cortex can function as units which have a much longer duration of electrical activity than individual cells, due to the ability of neurons to excite one another through recurrent loops (Amari, 1977; Lu et al. 1992). Thus the existence of these multicellular stages has an extension in time that individual cells lack. As a general rule, the larger the holon in the brain, the longer it may maintain a particular pattern of electrical activity. In this sense, it is both more stable and longer-living than its individual cells.

Finally, we note that cells within higher-order holons can have higher-order properties. That is, they can participate in the existence of higher dimensions, just as atoms in large molecular structures can. I discussed some examples of these properties earlier. Individual cells, when part of a tissue, can communicate with one another in ways that independent cells cannot. Chemical factors released or presented by one cell can regulate the growth or physiological activity of adjacent or even distant cells in very specific ways (Becker and Deamer 1991). Such regulation is not possible, or possible only to a very limited extent, by interacting cells outside of organisms. When cells are part of the brain, the highest and most developed organ, they may take on still even more sophisticated properties. Certain identifiable neurons in the visual cortex can respond to relatively complex stimuli, such as specific visual patterns (Wiesel and Hubel 1963; Crick and Koch, 1994). Their ability to do this depends on the existence of a complex network of connections with other neurons, and for just this reason, no cell existing outside of an organism could exhibit this kind of behavior.

Indeed, the process of perception offers one of the clearest examples of how emergent features of existence operate in different dimensions, and provides early hints of the most important and mysterious of all properties, consciousness. In Chapter 2 I suggested that we might define a level of existence in terms of six dimensions, three of space and three of time. In addition to functioning in a certain set of dimensions, a holon also has some kind of experience of these dimensions. For example, zero-dimensional experience sees the world as a point; that is, it views itself as everything there is. One-dimensional experience can make a distinction between self and other; it sees itself as a point, but has some knowledge of existence beyond this point.

Zero- and one-dimensional experience are typical of autonomous cells, that is, unicellular organisms. Cells within organisms, in contrast, may have an experience of higher dimensions10. This is particularly well-illustrated by cells in the visual pathway, where a holarchy of awareness can be observed (Baron 1987; Reid 1999). This pathway begins with the retina of the eye, where cells typically respond in a simple on or off fashion to flashes of light; that is, they increase their electrical actvitiy when a light either comes on or goes off, depending on the particular cell. This indiscriminate response to light is one-dimensional, or self-other, experience. Many neurons in the visual cortex of the brain, in contrast, can respond to lines of different orientations. So-called simple cells respond to a line of a characteristic angle, each cell responding to a line of a different angle. This is two-dimensional experience. Complex cells also respond to oriented lines, but unlike simple cells, can respond to a line present anywhere in the visual field. Thus they have some three-dimensional experience . Still other cells in the visual cortex can respond to lines that move in different directions. This kind of perception requires some awareness of time, so such cells exhibit some degree of four-dimensional experience11. As I pointed out earlier, another example of multi-dimensional experience is the ability of certain cells to learn, as this involves discrimination of a temporal dimension to a stimulus.

I want to reiterate that when I speak of dimensions, I'm not necessarily using them in the strict mathematical sense, in which a higher dimension has a relationship of infinity to a lower dimension. Dimensions as defined here emerge through the repetition of holons on one stage of existence, forming a new and higher stage. The reader should also remember that any set of dimensions is relative to one particular level. Though a cell may have zero-dimensional perception on the biological level, it's a six-dimensional holon (or at any rate, contains numerous dimensions) relative to its component atoms.

In the next chapter, we will see that organisms, too, differ in the number of dimensions in which they perceive the world, and this number is likewise related to the stage on their level of existence, the mental, with which they are associated. I regard this dimensionality of perception as extremely significant, because it suggests that holons on different levels of the holarchy are analogous with respect to not only what might be called their exterior properties--how they appear from the outside--but also their interior properties, what they actually experience (Wilber 1995; the distinction between exterior and interior will be discussed in more detail in the next chapter). Some theorists have protested that whatever laws or rules we determine from observing lower forms of life will be totally inapplicable to the latter kind of properties. "There is nothing in [lower level principles] that will tell us how to resolve an Oedipus complex, or why pride can be wounded, or what honor means, or why life is worth living," insists Ken Wilber.12 Likewise, in assessing the relevance of a model based on certain observations of physical systems to human societies, Rupert Sheldrake comments:

 

"A mathematical model of urbanization may shed light on the factors affecting the rate of urban growth, but it cannot account for the different architectural styles, cultures and religions found in, say, Brazilian and Indian cities."13

 

To the extent that architecture, culture and religion are interior properties--that is, an expression of the thoughts and feelings of people--we would not expect most models derived from physical or even biological systems to apply to them, because such models are usually based on exterior properties of holons. That is, when scientists study atoms, molecules and cells, they generally only observe their exterior properties. Therefore, any model based on such observations can only apply to exterior properties on higher levels, as Sheldrake suggests it might indeed do.

This does not mean, however, that there are no lower-level analogies applicable to our thoughts or feelings. It means that we would have to search for such analogs among the interior properties of lower levels--in other words, in the way these lower level holons perceive their world. The notion that perception can have differing degrees of dimension is, I think, a starting point for such a search; it is a way of getting at the question of the structure of the experience of lower forms of life. Further insight, perhaps, might come from understanding the quality of experience lower level holons have when they perceive their world14. This experience is obviously not the usual subject of scientific study, and perhaps is not even possible to access. However, it might be possible under certain conditions, for example, through the use of drugs that inactivated our higher level mental structures (see also the discussion in Chapter 7).

On the other hand, to the extent that human culture consists of exterior forms, a model based on lower forms of life might apply. Consider architectural styles, for example. A model based on cell phenomena might very well have some analogs at this level. If we haven't seen them, it's because we haven't looked for them. If the only thing that seems to be analogous in the particular model Sheldrake was discussing is urban growth, it's presumably because growth is all the model was meant to apply to, on any level. A study of physical or biological processes that focussed on diversity of shapes might well find some meaningful analogies between holons on these levels and architectural designs. Certainly the development of architecture has followed some evolutionary principles that apply to lower forms of life15.

 

Why Holarchy?

We have now seen that both cells and organisms are holarchically organized; they are holons containing several stages of lower-order holons within themselves. In the next two chapters, we will see that still higher levels of existence have a similar organization. Before we proceed further, though, we might ask, why is holarchical organization such a pervading theme in nature? Why are the same kinds of relationships found again and again?

The short answer is because holarchy is a very efficient means of organization; that is, it's a way of creating a tremendous amount of variety and novelty from a relatively few basic plans, structures or processes. Many biologists have pointed out that both molecular structures, within the cell, and multicellular arrays, within the organism, function to a great extent as interchangeable parts, or modules, which can be combined in many different ways (Raff 1996). "Relatively few medium-sized molecules are made by living cells,"16 notes Francis Crick, because they are require more synthetic steps than either small molecules, or large polymers.

Modern human technology uses the same principle. When a new car is designed, for example, the designer does not create an entirely new vehicle from scratch; rather, she combines established parts or themes--four wheels, front and back seats, head and tail lights, engine, transmission, and so on--in a new way. Just as there are a rather small number of different car designs, compared to the total number of cars, there are a relatively small number of body types, or bauplans (about 30) relatively to the immense number of organisms (Campbell et al. 1999). But an immense amount of variety is still possible within these basic plans. So while many questions remain about their origin (Mayr 1988), it's fairly clear that once they were established, it became easier for life to adapt by making modifications in them, rather than designing new ones from scratch.

Perhaps the most dramatic example of the great variety that can be generated from simplicity in this manner is the protein molecule. Proteins, as I discussed in Chapter 2, are composed of amino acids, of which there are about twenty in nature. Even a very small protein, containing just one hundred amino acids, can therefore exist in 20100 ~ 10130 varieties! Most of these possibilities are never created--indeed, this number is greater than the total of all atoms in the universe--but it illustrates the awesome amount of novelty that can be created from a few interchangeable parts.

Physical and biological holons, then, are something like the familiar children's lego set, in which a few building blocks are combined in diverse ways. But unlike lego units, which are simple physical structures, physical and biological holons, as we have seen, have dimensions in time as well as space. Furthermore, in addition to the simple contact interactions of lego blocks, physical and especially biological holons can be connected through a variety of different processes, again, over time as well as space. These added factors greatly multiply the variety that holarchical organization can produce.

Holarchical organization, in short, is a way of maximizing efficiency, of creating the most from the least. Indeed, I believe that efficiency, properly understood, is as good a candidate as any for a central organizing principle of the universe. It comes into play at all levels of existence. The physical chemist Manfred Eigen, whose hypercycles of mutually catalyzing metabolic reactions we will consider in the discussion of evolution in Part 2, notes that such self-organizing processes minimize energy usage by recycling the products of one reaction as starting materials of another (Eigen 1993; Lee 1997). The same principle was articulated by Alfred Lotka, a founding father of the science of ecology and the first scientist to treat populations and communities of organisms as thermodynamic systems:

 

"Evolution proceeds in such direction as to make the total energy flux through the system a maximum compatible with the constraints of the system."17

 

In other words, the system organizes itself so that everything is used, nothing is wasted.

Why is efficiency so critical? As I discussed earlier in this chapter, the organization of physical and biological holons is guided by transcription and translation of information contained in the genome. Transcription of the genome enables the cell to reproduce itself; translation of this information enables it to create a new organism. The latter process, in particular, poses a major problem. The amount of genetic information may seem vast, but it really is nowhere nearly adequate, by itself, to account for all the information that an organism effectively represents (Riedl 1978).

Consider: there are roughly one hundred thousand genes in the chromosomes of even complex organisms like ourselves, but billions of cells that must be arranged in a precise order. Obviously, if every cell-cell interaction had to be specified by even one gene, let alone many by genes, organisms in anything like we know them would not be possible. The amount of genetic material required would dwarf the resources of the cell. Organisms become possible because genes do not embody the detailed plan of the organism, but rather a few fundamental units--different kinds of cells--and a relatively few rules that determine how those parts will be assembled. Molecular biologists are gradually learning about these rules as they study how a few coding sequences of DNA, called regulatory genes, control the expression of many other genes during development (Dulbecco 1987; Kauffman 1993; Raff 1996; Gearhart and Kirschner 1999).

A fundamental advantage of holarchical structure, therefore, is that it can express, or unfold, a very large amount of information from a much smaller amount of information. Information that can be reduced in this manner is said to be compressible. The concept of compressibility was originally developed by the mathematician Gregory Chaitin (1988,1999) as a way of determining whether computer programs could be streamlined, that is, written with less code. Most computer programs proceed by making a series of logical, or "either-or" choices. As the mathematician Alan Turing showed, any program of this kind can therefore be represented by a string of digits consisting of just "0's" and "1's" (Hofstadter 1979). If the sequence of 0's and 1's is random--like that which would be generated by flipping a coin, for example, with heads considered 1 and tails 0--there is no way that the information it contains can be conveyed in any message shorter than the entire string itself. Some strings, however, have certain patterns, such as the one that goes 01010101...All the information present in this string can be conveyed by a shorter string or program that says, in effect, "repeat "01" n times". The information in such a string or program is therefore said to be compressible.

Chaitin's theory (called algorithmic information theory, or AIT) was originally used to demonstrate that randomness is a problem in a wider variety of mathematical operations than was previously appreciated. But it may have important implications for our understanding of living things, and particularly in the holarchical model of existence. As I noted above, it's clear that the information present in an organism is compressed in its genome. While all the information in an organism could be stored as a complex description of every cell's interaction with every other cell, the genome compresses this information into a few thousand sequences of nucleotide bases, and perhaps a few dozen or hundred rules governing when, where and how these genes are to be expressed. This compressibility, much like that of Chaitin's strings, results from the highly repetitive, patterned nature of life that is the hallmark of holarchy. Thus protein molecules are composed of repeating units of amino acids; tissues are composed of repeating units of cells. The information needed to specify such patterned interactions can obviously be compressed greatly from a program that specified each and every interaction.

The compressibility of the information represented in the whole organism is therefore evidence of the efficiency of holarchical development. But what about the information in the genome itself? Could that be further compressed? In other words, could we in principle design a set of genes that could generate a real organism, using fewer nucleotide bases than are actually present in the genome of that organism? The available evidence suggests we probably could. We know, to begin with, that most of the DNA in the genome does not code for any protein. Most coding sequences are broken up by non-coding sequences, sometimes immensely long, called introns. In most cases these introns have no known function, though some scientists speculate that they may play a role in creating chromosomal structure (Brown 1999). Their presence also makes it possible for different coding sequences to combine into new kinds of proteins, which may have been vital to the early evolution of variety (Gilbert 1987; Gilbert and Glynias 1993). Very recently, it has been suggested that introns may contain information in the same sense that coding sequences (exons) do (Flam 1994; Moore 1996).

Another observation strongly suggesting that the information in the genome is not completely compressed is the apparent existence of redundancy of genes. If every gene expressed a unique and critical piece of information for the organism, we might expect that mutation or elimination of that gene would result in the death of the organism, or at least in a major change in some function. Studies have shown, however, that not all genes are critical in this sense. While the exact proportion of vital genes in human is not known, it has been estimated to be around 40% (Lewin 1997). This doesn't mean that the other 60% of the genes could be eliminated en masse, but it does indicate that many genes don't do anything that is not done by some other gene.

The lack of complete compressibility of the genome, in the view of many scientists, is a powerful argument that the genome, at least in its modern form, is a product of Darwinian evolution. I will discuss evolution in detail in Part 2, but here I want to point out that because Darwinism does have a strong element of chance, it may create forms of existence that are not the most efficient, at least from our point of view. Furthermore, because this kind of evolution frequently, though perhaps not always, proceeds in small steps, it becomes committed to certain directions; as life becomes more complex, it becomes more difficult for it to back up, so to speak, to move in a new direction.

 

Information, Energy and Complexity

Another very significant implication of Chaitin's work is that it may enable us to integrate three key concepts that appear again and again in discussions of holarchical relationships: complexity, information and energy. Most holarchical theorists assume that higher forms of life are more complex than lower. Thus a cell is said to be more complex than an atom or molecule, and a multicellular organism more complex than a cell. Complexity, as applied to holons in this way, is virtually synonomous with "level (or stage) of existence". We say that one form of life is more complex than another when it is higher than the latter, and includes the lower within it. The higher is by definition more complex than the lower.

We have also seen that energy, too, is closely associated with holarchical status. As I noted earlier, higher-order holons contain more energy than lower ones. Thus energy is required to synthesize larger molecules (such as proteins) from smaller ones (such as amino acids), while energy is provided by breaking down larger molecules (such as sugar) into smaller ones (carbon dioxide and water). Likewise, energy is required for creating tissues and other multicellular arrays from cells.

More specifically, energy seems to be synonomous with the hetarchical interactions formed between fundamental holons. These interactions can be said to store the energy, or to make it manifest in a structural form. At its most fundamental physical level (or subphysical in this model of the holarchy), energy is embodied in the interactions of subatomic particles. Within the cell energy is represented by chemical bonds between atoms; in the organism, by cellular interactions in tissues and organs. On a still higher level, as we shall see in the following chapter, energy is embodied in the interactions between organisms, and particularly between human beings.

Into this mix, Chaitin brings his concept of information. We have already seen that in AIT, a noncompressible string or program is considered to have more information than a compressible one of the same length. This information, in AIT, is directly related to complexity. Thus Chaitin defines the complexity of a binary string as "the minimum quantity of information needed to define the string."18. So in a general sense, complexity, energy, and information seem to parallel each other, each increasing as the holarchy is ascended. These are all candidate concepts, along with that of freedom that I introduced in Chapter 2, by which to define, and perhaps even measure, what it means to be higher in the holarchy.

However, the relationship is almost certainly more complicated than this. Information, in Chaitin's terms, is associated with randomness. A completely random sequence of digits--or of anything else--is considered to have more information, and more complexity--than a patterned one of the same length. This seems counter-intuitive. When we see a random string of digits, we think of it as having no information at all, whereas a highly patterned string does seem to have information. Or, to put the notion in more famiiar terms, when we speak or write language, which is based on certain rules which give our words some pattern, we transmit information, whereas there seems to be no information at all in a purely random sequence of words or letters (gibberish). Likewise, randomness in physics is conventionally associated with entropy, a lack of complexity or order. A random mixture of two gasses such as nitrogen and oxygen, for example, has more entropy than two, separated quantitites of pure oxygen and pure nitrogen.

The reason for this apparent discrepancy between our common-sense (and indeed, conventional scientific) understanding of randomness, on the one hand, and Chaitin's definition, on the other, is that Chaitin is considering one particular random state, while we are considering all possible states. Consider again a random sequence of words. What we normally mean by this notion is any random sequence of words--that is, any one of a very large number of sequences, each of which is random. By Chaitin's definition, the information involved in expressing this notion is indeed highly compressible. That is, a much shorter string or program could be written to express the notion of any random sequence. The program would simply say: generate all possible strings of such-and-such a length. So randomness in this sense does have low informational content, and low complexity, in Chaitin's AIT.

Chaitin, however, is primarily concerned with the informational content of one particular random sequence of words. This is not really "gibberish" or "nonsense" because to pick out one particular such sequence out of the astronomically large number of possible random sequences is not what we normally call a random process. In order to define such a sequence, a great deal of information would be necessary.

Nevertheless, some information theorists have argued that complexity is not a simple function of randomness, even understood in this manner. Charles Bennett has attempted to define complexity in terms of something he calls "logical depth" (Davies 1992; Norretranders 1998). In Bennett's view, complexity is directly related to the degree of neither randomness nor of order, but emerges in an area somewhere between them.

Perhaps the best example of complexity in this sense is provided by language. If our language were perfectly ordered, it would consist of say, a single letter, repeated a number of times. If it were perfectly disordered, it would consist of every possible combination of letters. Neither type of language, obviously, is practical. A language with high order doesn't have enough information; if all words are spelled with the same letter, we can't create enough words. A language with high disorder, however, would be too difficult to use. For one thing, many words would be unpronounceable; for another, there would be more words than anyone could possible remember. So human language can be thought of as a kind of compromise between randomness and order. Words can be composed of many different combinations of letters, but there are some rules, which provide a degree of order. Such a compromise, Bennett believes, is what true complexity is all about.

Another example of such a compromise, on a lower level of existence, is provided by protein molecules. If proteins were highly ordered, they would all consist of one kind of amino acid; if they were completely random, they would consist of any possible combination of the twenty or so amino acids found in nature. Again, both ends of this spectrum are impractical. There is not enough information in proteins composed of just one amino acid, and there is too much information in proteins that are completely random. Real proteins are composed of many different kinds of amino acids, but they are probably not completely random. The must fold up into highly specific three-dimensional shapes to carry out their functions, and the interactions underlying such folding put some constraints on the kinds of amino acids that can be present in particular positions of the sequence. Most proteins that have been identified fall into one of probably no more than a few thousand "superfamilies", members of which share many amino acids in common (Creighton 1993).

Bennett's view of complexity is thus a common-sense one; it seems to reflect what we intuitively mean when we say something is complex. Roughly speaking, complexity corresponds to what we call "useful information", while the extra information in his sense we would call "junk". Another way of putting it is to say that complexity is information constrained by order.

Like many definitions based on common sense, though, this one runs into problems when it's treated rigorously. The only way that Bennett has come up with to quantitate his complexity is by the amount of time or effort that it takes to compress or filter information--to take out all the "junk", so to speak, and order the remainder into useful information. For example, one would measure the complexity of a book not by counting the bits of information it contained, but by the time or effort it took the author to write it. If complexity is defined in terms of time or effort, however, it can only be measured empirically, that is, by observing the information-containing entity as it actually comes into existence. This may be possible for a book, but what about an evolving cell or organism? How are we going to measure its complexity?

Even in the case of the book, a problem arises in that there is always the question of whether the time or effort that it took a certain pattern of information to be created was the least possible. Two people might be capable of writing the same book using greatly different amounts of time and effort. Even if we somehow defined a "standard" author, we could never be sure that she couldn't have written the book faster or more easily, using some different approach.

One very important idea that has emerged from the work of Bennett and others, though, is that the discarding or erasing of information may be as critical to the understanding of complexity as information acquisition. If complexity lies between maximum information and maximum order, then to convert information to complexity, it seems that we have to throw out a lot of junk. The Danish writer Tor Norretranders, whose book The User Illusion (1998) explores this idea in detail, notes that we human information processors are conscious of but a vanishly small fraction of all the information that impinges on our senses. Our mental activity thus results from what Norretranders calls exformation, a process of discarding information. Extrapolating to other levels of existence, we might argue that a similar problem faced lower forms of life.

I will discuss the concept of information further in Chapter 8, when we examine evolution; and I will return to the subject of consciousness in Chapters 4 and 5. For now, I want to emphasize that information, as it becomes better defined, is poised to play an extremely powerful role in our understanding of life. Some theorists have suggested that information may be a fundamental feature of the universe (Chalmers 1996; Davies 1999), and that a better grasp of it could help us fill in the critical gaps in our understanding of how evolution occurred. Our ability to measure, in a meaningful way, the amount of information in the genome seems a very reasonable possibility following the total sequencing of it, which will be completed in a few years (Cooper 1994; Norman 1999; Haddad et al. 1999). This may allow us to understand evolution in a new and powerful way.

Go to Footnotes

Go to Chapter Summary

Related articles: