From Section 4: Patterns of Change

Throughout this chapter we have recognized many different kinds of living things as persistent patterns which emerge from interactions involving structural components (i.e. atoms, molecules and macromolecular complexes), living constituents (i.e. cells, organisms, communities) and various aspects of the physical environment. Those patterns can embrace many different scales of time, space and organization, and their development and survival can involve interactions as variable as the sprouting of new plant communities, as precise as the processing of macromolecular information, as mysterious as the epigenesis of complex organisms and as subtle as the relationships that maintain the non-equilibrium sort-of-steady state that supports all life on Earth. Yet no matter how large or long-lived or complex they are, all of the living things we have encountered so far are susceptible to change and death, and as a consequence each living thing is also unique, whether we compare it to others of its own kind or to itself at different points in time. Living communities embrace unique and dynamic networks of constituents which interact with environments that are altered by their activities and subject to the influence of geophysical factors that can produce everything from subtle local and seasonal variations to major rearrangements and global catastrophes. Within those communities organisms are continually changing as they grow, regenerate, degenerate and die, and the persistence of each kind depends upon the continual contribution of offspring to a vast and ever-changing variety of lineages. Within those lineages sophisticated information-handling abilities allow some organisms to pass their patterns on with considerable precision, but the uniqueness of each individual is ensured by the physical impossibility of exact reproduction, even when it is desirable (more on this in a moment). As for the living Earth as a whole, we can hardly expect the biosphere to remain homogeneous or static when its inhabitants and their natural situations have not, and a substantial body of evidence tells us that our unique and ever-changing planet has indeed been host to a rich and dynamic parade of creatures and communities, which have seldom survived or remained the same for long, at least not long by geophysical standards.

So, if there is one basic conclusion we can draw from what we have learned so far, it is that wherever and whenever we look in the living world we can expect to find change and variety. This may seem a rather modest destination to arrive at after such a lengthy journey, but as has been mentioned at several points along the way, such a conclusion is not obvious to people who are used to looking at the world from an idealistic perspective. Take the naturalists who toiled through the generations separating us from Aristotle, for example. Many of them considered earthly objects as diverse as geometric shapes, rocks and living things to be reflections of universal and unchanging ideal forms, and the differences they observed among such objects were either classed as essential traits (i.e. differences that distinguish one species from another), or dismissed as inconsequential variations within species.

A world view that denies change and dismisses variety can be quite useful if your aim is to systematize knowledge as Aristotle did. A sense of ideal order also comes in handy for explicitly logical endeavours like geometry and mathematics, and in the study of things that appear to behave logically, like steam engines, atoms and angels. However, as naturalists expanded their exploration of the living world they found it increasingly difficult to make sense of what was going on around them and what appears to have happened in the past, and at the same time remain faithful to the idealistic tradition. Those who tried to do so were obliged to resort to increasingly dubious measures, such as packing infinite generations of homunculi inside eggs and sperm or invoking invisible forces to control the survival and development of everything from cells to communities. But the naturalists who were willing to recognize change and variety as fundamental aspects of the living world were able to assemble a rational framework that may have lost some ancient certainties, but it has proven immensely useful in efforts to understand what is going on. For instance it leaves us free to ask questions that went unasked for a long time, such as: how and why do things change within the living world?

Process and Progress

Once again the thread leads us back to Aristotle, who as we saw earlier considered the development of organisms to involve the self-realisation of ideal forms that are inherent in various substances, such as the combinations of fluids that give rise to animals. This is sometimes referred to as order arising out of chaos, but it is important to note that according to Aristotle order is never created, destroyed or changed, it is simply manifested in a visible material form as substances evolve towards the complete expression, or perfection, of their inherent order. Thus in the Aristotelian scheme each evolutionary journey - whether of an organism or the entire universe - is a process, or a sequence of events leading to a specific goal, which is driven by the innate desire of form to realize itself. In other words, form naturally seeks "mastery" over matter. Circumstances can influence how rapidly particular evolutionary journeys progress; for example seeds that fall on barren ground sprout more slowly than those planted in rich soil. Circumstances also influence how far such journeys proceed; for example some seedlings are trampled before they can become trees. But according to Aristotle, nothing can change the ultimate goal of a journey or the route that must be followed to achieve it. For example a tree seedling that sprouts from barren ground still struggles to grow into a tree rather than a blade of grass.

Like most of Aristotle's ideas, the notion that evolution is a process that progresses towards perfection has a strong commonsense appeal. For example the development of an egg into a chick certainly looks like a process with a well-defined goal, and most people would consider a chick to be a more complete expression of the ideal of "chicken-ness" than an egg. The concepts of process and progress can be easily applied to many of the changes we see going on in the world around us, and these concepts have also proven useful in efforts to delve deeper into the details of how and why things change the way they do, in some cases supporting propositions that contradict Aristotle's teachings. For instance he denied the existence of matterless form and had little use for divine forces, yet philosophers who claimed to have rediscovered Aristotelian principles had no qualms about populating the universe with a host of forms that either transiently animate bodies, such as immortal souls, or remain permanently unincorporated while they guide earthly events in accordance with the dictates of the divine intellect, such as Thomas Aquinas' angels. Those angels faded into the background during the Renaissance, when many Western thinkers adopted the more classically Aristotelian view that the world was created in such a way that it naturally progresses towards perfection. This deist viewpoint allowed the world to evolve while accepting Aristotle's teaching that species do not, yet Lamarck was able to take a tempting half-step away from this position when he proposed that organism lineages can change, albeit in a progressive manner. He attributed the evolution of lineages to the active striving of individuals to adapt to their surroundings, which leads them to pass acquired modifications on to their offspring. Thus Lamarckian evolution is a process that is driven by the desire of each creature to become perfectly adapted to its natural situation.

The concept of evolution as a progressive process is still very much with us today - indeed process and progress have become virtually intuitive concepts in the modern mind. For example it is common practice to rank individuals, organizations, nations and even larger aggregations according to how far they have progressed along the universal evolutionary route to "development" - for instance is a "developed" world and a "developing" one, and how committed they are to following that path (i.e. how "progressive" they are). Nevertheless, I trust that most readers have noticed that process and progress do not figure prominently among the planks of the conceptual framework we have been assembling here; indeed I have gone to considerable lengths to exclude them.

This was fairly easy to do when we were exploring the development of natural communities. For example the tropical rainforests that have appeared in different parts of the world may have the same general structure and they may evolve in similar ways, but a close examination does not support the notion that they are the explicit products of a common developmental process - the same consideration applies to forests that "regenerate" in cleared areas. A simpler explanation for the situations we observe is that similar communities emerge when similar combinations of creatures interact with each other in similar ways in similar environments under similar conditions - in other words the similarities we observe among communities are the convergent outcomes of unique evolutionary journeys, and by the same reasoning their differences arise from evolutionary divergences. It may have been less obvious, but we arrived at essentially the same conclusion when we examined the development of complex organisms and decided that the simplest way to account for the observation that closely related individuals turn out alike is that similar patterns emerge when similar components and constituents processing similar information in similar ways interact with each other under similar circumstances.

It is important to note that while the concept of evolution that emerges from our picture of the living world does not involve goals, absolute order or progress towards perfection, this does not mean that we have left everything to chance as living things wander randomly through time - indeed it is only after we free them from absolute control and the obligation to progress that we can begin to understand the influences that shape their evolutionary journeys. For example many plant communities continually receive seeds that drift in randomly on the wind, but the success of the new arrivals still depends upon the opportunities that are available, with the seeds of sun-loving shrubs being as unlikely to fare well on the shady floors of mature hardwood forests as the seeds of mighty trees in an open patch of meadow. Chance is also involved in the coming together of gametes to form a zygote, although of course the developmental potential of each zygote is primarily defined by its ancestry; for example those with horse genomes become horses, those with donkey genomes become donkeys, and those with hybrid horse/donkey genomes become mules or hinnies. Chance exposure to various factors and experiences can influence the evolution of individual cell lineages within organisms, but their evolution is far from random because cells respond to specific aspects of their surroundings in specific ways, such as by activating the expression of particular genes. Cell lineages also tend to follow consistent developmental patterns owing to the way they process and accumulate information. For example by the time erythrocyte precursors appear in a blood cell lineage the only major developmental decision they can make is to either complete development or die; in order to explore any other possibilities the cells would have to unmake decisions that accumulated over several generations, which is not possible because the outcomes of many of those decisions are embodied in the cells' physical patterns.

Thus in place of a process that progresses towards perfection, our conception of evolution centres on the unique journeys living things follow as they are influenced by interactions with their surroundings, by interactions among their components and constituents, by the information they have to work with, by their own past and by the past histories of any lineages they may belong to. Those journeys are not random, but they show no indications of having predetermined goals or routes - since evolution is not a process - and they are not progressive in the classical sense because they do not involve the realisation of absolute order, but rather the ongoing creation, alteration and destruction of dynamic order. This leads us to a vision of the living world where everything can change and nothing is obliged to progress towards perfection, with perfection itself becoming a dubious notion in a world where everything can change. For example many of the creatures and communities that vanished during past mass extinction events may have been perfectly suited to the prevailing environmental conditions, but that did them little good when those conditions changed.

The more attentive readers will have noticed that the preceding paragraph contains a subtle shift from the concept of evolution as something that involves individual living things - or what is commonly referred to as development - to the concept of the evolution of kinds of living things, such as lineages of organisms, which is what most people think of when they encounter the term "evolution." Actually, according to our conceptual framework there is no difference between the evolution of individual organisms and the evolution of their lineages, since neither involve processes or progress and both are shaped by the same influences: interactions, information, history and chance. I expect that this would be obvious to most readers by now if it was not for the fact that in the preceding section I emphasised how effective cells and organisms are at conserving their hereditary information. If all genomes were perfectly preserved and passed on we could hardly expect to see hereditary lineages change over time or diverge from each other, but since they do evolve we have good reason to suspect that genomic information is not perfectly conserved. As it turns out, not only is conservation impossible, in many cases it does not even seem to be desirable.

Inevitable Variations and Deliberate Rearrangements

I mentioned earlier that genomic information is highly conserved within the cell lineages that arise from the same founder cell during epigenesis, and provided some adjective-adverse editor did not delete it the word highly was important, because it is a fact of cellular life that hereditary information cannot be absolutely conserved, even when the cellular machinery is dedicated to doing so. We can begin to appreciate why this is so by considering the magnitude of the challenge a cell faces in passing complete and intact copies of its genome to its daughter cells during division. For example every time a typical human cell divides, 46 large chromosomes must be unpacked so that their DNA can be replicated - a total of 5 x 10⁹ base pairs in all - and then the DNA strands have to be repacked into chromosomes and distributed so that each daughter cell gets a complete set.

This is a formidable challenge; indeed as far as I know, humankind has yet to come up with a process that can consistently replicate information with an error rate of less than one in a half billion. Nature has, of course, had a lot longer to deal with this challenge than we have, and studies have shown that on average the replication complexes in human cells add one wrong nucleotide base for every 10 million they copy. This is an impressive level of accuracy, and as human engineers have discovered during their efforts to build information processing systems, the accuracy can be further increased by proof-reading mechanisms. Such mechanisms have been incorporated into the cellular replication machinery, where they have reduced the error rate to approximately one permanent base change (i.e. point mutation) for every 10¹⁰ DNA base pairs copied. This means that on average the machinery creates one new point mutation for every two complete human genome replications of 5 x 10⁹ bases each, making the odds about 50:50 that a daughter cell will inherit a new point mutation when a cell divides.

Those are pretty good odds when we consider the likelihood of an individual cell lineage keeping its genome relatively unchanged, but the situation changes somewhat when we shift our perspective to a complete organism. For example the typical lifetime of a human body involves over 10¹⁵ cell divisions, which means that in order to perfectly conserve the zygote genome, more than 10²⁴ DNA base pairs must be replicated without error, and almost 10¹⁶ chromosomes must sort themselves correctly among daughter cells. Needless to say, this is impossible. In fact, if we assume that on average the replication machinery creates a new point mutation within each cell lineage every two generations, then it follows that if we could track the descendants of any particular human zygote we would expect to see every single base pair in the original genome mutate about fifty thousand times (keeners can do the math for themselves). Of course the odds of a finding a particular mutation in a particular base pair in a particular sequence in a particular cell would still be on the order of one in a hundred million.

Another factor that must be take into consideration in the challenge of conserving hereditary information is the fact that the information comes in the form of large, complex macromolecules, which can serve as targets for a wide range of disruptive influences. For example DNA strands can be nicked, broken or chemically altered by energetic ions and ultraviolet radiation. The significance of this threat is shown by the fact that cells have come up with a variety of ways of protecting DNA from damage, with the first line of defence being the chromosomes, where DNA macromolecules are wound up with proteins to form compact and chemically stable structures. Chromosomes are highly stable but the DNA inside them is not immune to damage, and it also has to be unravelled occasionally to be replicated and to allow genes to be transcribed. This vulnerability has obliged cells to come up with a second line of defence against DNA damage, in the form of machinery that can detect and fix defects like strand breaks, gaps in nucleotide base sequences, chemically damaged bases and mismatches where non-complementary bases end up across from each other (e.g. a guanine on one strand faces something other than a cytosine on the opposite strand). Most cells possess machinery that is highly efficient at detecting and fixing such mutations, but its effectiveness is limited by the fact that the repair machinery is not very good at interpreting the information it protects. For example the repair machinery of bacterial and eukaryotic cells can easily detect nucleotide base mismatches between DNA strands, but the machinery has no way of telling which strand has acquired the "wrong" base relative to the sequence that existed before the mismatch arose. As a result the repair machinery arbitrarily cuts out one of the mismatched bases and replaces it with one that fits, which means that on average half of all repaired sequences acquire a new point mutation.

These basic considerations and simple calculations are sufficient to prove that it is virtually impossible for an organism as large and complex as a human body to be made of cells with identical genomes, and we can apply similar considerations to the passage of hereditary information from generation to generation. For example the germ cell lineages that give rise to human eggs only go through a few divisions in the fetus, which means that on average each egg carries one or two new point mutations (although eggs can get into other kinds of trouble, as we will soon see). In contrast, the germ cells that continually give rise to sperm throughout the lifetime of human males go through several divisions before they reach their mature form, by which time the odds are that each sperm genome will have accumulated about a hundred point mutations. This means that every human zygote starts out with a hundred or so new genomic mutations courtesy of its father, and it also means that in principle the number of sperm required to collectively contain a new point mutation in every base pair in the human genome would be around 100 million - which by an interesting coincidence is about how many cells there are in a typical ejaculation......