Big Black Box Still Remains

Discussion of the nature of emergent traits draws attention to at least one box that remains rather completely black. Inside that black box is perception. Perception is made possible by patterning processes, activation/inhibition, cell adhesion, ligand binding to alter gene expression, and so on—simple developmental processes that in natural and understandable ways provide spatiotemporal maps from sense to sensor. The developmental processes generating sensoritopic maps account for how individual incoming signal properties, like wavelength, sound frequency, or specific odorant molecules, can (in some species and for some senses) be distinct in an orderly and sufficiently replicable way. This is what we have seen for other organ systems.

Perhaps the most interesting aspect of sensory and cognitive systems, however, is not their molecular components but what these perceptive systems "feel" like to the organism, their emergent property. For example, how and why does hearing feel different from seeing when both involve structurally and chemically very similar neural systems (indeed, that can remap from one type of sensation to another)? This immediately leads to the issue of most interest to humans: the relationships between perception and consciousness.

As far as keeping track of incoming signals, one might liken the brain to a building guard sitting in the guardroom of a bank, watching the halls, teller windows, and so on through television monitors (it is in fact debated, but immaterial to our point here, whether the brain actually works by having a monitoring center— sometimes also referred to as a "homunculus"—or not). The guard does not see anything directly and must respond on the basis of video images provided by the monitors. This is inherently a limited amount of information compared with what is really going on out there in the bank. However, because television monitors are designed specifically to mimic human visual perceptive experience, the guard does have a rather natural, if one step removed, sense of at least that slice of external reality.

A somewhat more apt image might be to submerge the monitor more deeply, and think of a sailor in a submarine working with sonar and radio information. Sonar does not represent a natural human information gathering system. Why? Because undersea sounds are not the kind we are used to or evolved for interpreting. Undersea sound has to be translated into something interpretable, and we choose mainly to do that on a video screen, as blips identified as to distance, direction, and azimuth (or as audible beeps produced in earphones). These are all arbitrary ways to represent the information. Electronic detection (e.g., of radio signals) may be presented in some other computer-digested form. Indeed, sensors are made to detect things humans have no sensory means to detect (e.g., very low frequency radiation), and this information may be presented as spectral frequency pattern data, such as on an oscilloscope. In some cases information is presented in the even more abstract and arbitrary form as text on a screen, representing a totally black box analysis by a computer. Again, the data are translated into a form interpretable by humans, and the person has essentially only limited data, indirect contact, and must rely on integrated experience to interpret what the sensors provide. Every organism lives in a Kantian world, and can only know aspects of reality for which it has some form of receptors.

In this imagery, what counts is that some aspect of the signal, which we select for our own purposes such as to thwart burglars or survive attack, is chosen for detection. We design our interpretive machinery from the outside, perhaps using cost or other considerations to decide what aspects of the signal to bother about. Evolution has had to do this by trial and error from the inside, and the only known criterion is reproductive success. Each natural sensory system is built today to detect a subset of environmental information in a way that was evolutionarily sufficient in the past. But the neurons themselves, like the electrons used in the submarine's systems, are not directly related to the nature of what is being detected by them.

It is not surprising that in many systems the topographic relationship between the signal and the interpretation is maintained. This helps the brain keep track of important aspects of the signal, although what counts as important varies. Above all, it seems to mean replication: the same signal will fire the same distinct set of neurons. In some systems like visual image perception and touch, the map directly reflects the structure of the outside world, while for others like olfaction it seems mainly to serve a bookkeeping function. Vertebrate hearing uses a physical trick of forming a fluid-filled tube to decompose a complex wavelike signal. None of these particular kinds of sensory maps are essential, however; some species ignore aspects of these information sources while detecting others.

We have not really considered other "higher" aspects of the integration of environmental information, like prey tracking, locomotion, mating, eating, and the like, which integrate environmental sensing, physical activity, proprioception, and so on. These are internal supersystems, integrating muscular, neural, sensory, and motor functions, each of which is an organized system in itself. Emergence upon emergence. The overall ways these complex integrative functions are orchestrated are not understood (although some of the wiring is known).

We have suggested that these functions, including consciousness itself, evolved gradually and may constitute traits of which we have not much more direct sense of the reality than a sonar operator does of things detected from the murky deep. But this intriguing subject has to remain for the future. Even attempts to specify what the phenomenon of consciousness actually is are confessedly highly speculative (Crick and Koch 2003). Many biologists would argue that, with present knowledge and tools, these questions are currently beyond the reach of science.

some cautions about evolutionary and genetic interpretations

One problem in reconstructing life is that adaptation—whatever its cause—is not perfect, and there is no way to define what "perfect" might mean. If the only criterion is that the fit of the organism to the environment be "good enough," and this depends on the changing landscape of local environments, competitors, colleagues, and the genotypes available, then we cannot expect definitive answers. Rather, the generalization seems to be that there is no one way and no need to be better than good enough under the circumstances—and lucky.

Members of some species are largely safe from predation (elephants, lions, humans, giant turtles, probably many viruses and bacteria, and so on). Some seem not just essentially safe but to have rather open life spans, such as venerable olives, the famous Tule tree in Oaxaca, Mexico, and some giant sequoias in California that are thousands of years old. Other species live but a fleeting hour or so before reproducing (e.g., bacteria, adult mayflies). Some produce young that are almost all immediately devoured. Every creature gets through life differently.

Interestingly, however, each must face the same kinds of challenges, of escaping from predators and microbial attack and finding food sources. The differences are mostly of scale and circumstance. There is nothing inherent in mammaldom that should make them vulnerable to a greater diversity of parasites, for example, than a tree that has to just stand there and take it. This means that there is little in the way of a priori criteria by which to predict how or what a given organism will do or be like. It is in this important sense that the organism-as-machine metaphor can be highly misleading; it is thinking of organisms in human terms, as being designed to solve problems laid before them by nature. An individual organism may try to get "there" from "here" (e.g., to catch that rabbit), but evolutionarily the only problem it is solving is to persist and reproduce (to catch a rabbit today). And this is why it can be misleading to think of there being a fixed number of senses, or ways to hunt, or a single kind of "immune" system, and so on.

Not only is chance a pervasive factor in the environments inside and outside of organisms, but imperfection, if we can use the term, is an essential factor that makes evolution possible and ongoing. Energetic inefficiency is what enables evolution and in that sense it is non-sense to ask whether a trait has evolved to be energetically efficient. The puffer fish makes do with a genome only about 1/8th the size of its vertebrate relatives who, nonetheless, have roughly the same genes. This shows that such baggage can in practice be off-loaded to save on the substantial metabolic demands it must make. Yet selection is often invoked to account even for what seem to be the fine points of nonfunctional DNA (for example, it is said that there can be no truly nonfunctional DNA or selection would have eliminated it). But if energetic considerations of that kind applied as often as they are invoked, most individuals in most species would be on the brink of starvation, and thus need to shed every ounce of needless base pairs. In energetic terms themselves, purging has clearly not been worth the cost. If selection were typically too stringent, evolution as we know it might not have been possible—nobody would have survived.

Again, it is necessary to think contingently and that means separately for each case, and one result is a substitution of description for scientific generality. But that itself seems to be one of the realities. For example, one can ask whether evolution has made some trait more efficient than it used to be for the same use, but the truth is only that the trait is as efficient as it has had to be and the uses are always changing. A classic example is the argument that mammalian legs are oriented to work more for-and-aft than the arc-sweeping of reptile limbs, and that this evolved because it produced more efficient locomotion in mammals. But think of alligators in motion. Whether they use more energy per foot traveled than mammals, many a mammal has paid the ultimate price despite having "more efficient" locomotion. Or are the winners among competing alligators those with more efficient inefficient locomotion?

This exemplifies the problem of making evolutionary reconstruction stories based on how any particular trait "must" have come about because the post hoc nature of evolutionary inference means that multiple explanations can have comparable plausibility. This is known as a nonidentifiability problem, and it is a ubiquitous fact of evolutionary biology.

This is relevant to the persistent division of opinion about the roles of competition and cooperation in evolution. If a genetic variant becomes more common over time, one can always assume or define its increased frequency as due to adaptational competition and can in principle then look back into the specific history of the gene(s) and infer that what happened was deterministic. From this point of view, persons who see (or desire to see) cooperation as being as important in nature as competition can always be refuted: if social organisms, organelles, genes, or individuals cooperate and this leads to differential proliferation of associated genetic variation, then that cooperation can be expressed in a consistent way in terms of competition (some alleles do win, after all, if that's how we define winning).

But if the World According to Hobbes is a consistent one, the same outcome— what we observe today—can be accounted in a different way. Chance and organismal selection can lead to genetic change without the kind of darwinian warfare that makes for good television viewing (and has so often been used to justify social inequity). Organisms build their own environments and choose their own niches when they can. If it looks like cooperation, or feels like cooperation, then for all practical purposes it is cooperation on the level of organization at which it actually occurs. Cooperation is fundamental to organized life, from genes on up.

Focusing on how or why cooperation, or cultural inheritance, or social behavior, or organismal selection are "really" just wolves in sheep's clothing may help explain some aspects of life; however, this focus can lead to tunnel vision, drawing attention away from important or even pervasive aspects of how life works. If evolution is the meandering contingent process that it seems to be, the ultimate explanation that cooperation really represents a past history of competition may be true but too generic to explain very much. It is true that our house is made of nails, paint, and boards, but that does not explain our house.

In many ways, the competition-cooperation distinction is one more instance of a false distinction of perspective. Sequestration by itself almost implies cooperation if complex organization is to evolve because the interaction of isolated units like cells is the essence of such complexity. Without "cooperation" between various biological molecules, nothing in life happens. The catalyzing of biological reactions was for much of the 20th century taken to be a definition of the difference between life and nonlife. DNA does not even replicate itself without help, despite that often being stated as a biological fundamental.

Chance is always present as is the potential for natural selection, but the same can be said of organismal selection, a potentially less combative source of adaptation. Organisms sort themselves into local environments depending on what their capabilities are, and over time genetic changes that need not have to do with classical adaptation (e.g., chromosome rearrangements) can produce a species barrier. Indeed, that they are mobile and seeking is one of the traditional definitions of what it means to be an "animal." Facultative searching of environments is also done by single-celled organisms and even by plants.

Like sexual selection, self-sorting by organismal selection can be faster and more precisely related to function than classical natural selection, because organisms know better what they can do best than the crude screen of selection may be able to detect. There is nothing unnatural about this form of sorting and proliferating of genetic and phenotypic variation, perhaps supplemented by genetic assimilation, though it has been treated as a kind of backwater of biological theory. Instead, the search for such mechanisms as a way to account for adaptations should be more active.

Not Laws, But Principles, and Why There is no Argus

We have identified a variety of principles that we believe apply generally to evolution as it has occurred, and we think these make a useful addition to the usual darwinian principles. They are not new to us, nor to biological thinking, but they are not usually treated as part of the theory of evolution itself. If our view is justified, we might ask if these principles have any predictive power, one of the criteria for generalizations in science. In fact, we think this is indeed the case.

The pattern of the traits that we have described in this book is largely a very orderly one. Traits evolve differences, and new traits arise, out of earlier stages and by reusing processes that already exist. Traits of more ancient origin are more persistent, and derived traits are constrained by that fact. These general statements are true at the gene, morphology, and biochemical levels. As a result we can explain in a fairly formal and rigorous way, why certain patterns occur and others do not. There is much flexibility in these constraints as shown, for example, by the intercalative nature of the apparent re-evolution of traits like eyes. But because of the constraints we have described, we would agree with another observation of Thomas Browne (whom we quoted in Chapter 14 in regard to eyes and ^gles (Browne 1646)), that in "sanguineous" (vertebrate) animals, there are but two eyes, and they are in the head. There can be no Argus. In 1646 he was surmising, but evolutionary biology can explain the reason why.

There are undoubtedly errors in this book, although we hope they are not too many. Errors arise first from our own misunderstanding of existing knowledge in areas we have tried to represent that are beyond our own prior expertise, and then from the incompleteness and rapidly changing nature of that knowledge even if we have interpreted it accurately. However, the broad picture presented in these pages is likely to be robust even in the face of those kinds of error. It is always possible that fundamental new properties of genetic life will be discovered, and new fossil or living species are sure to be found.We cannot know what they might be, but from what we do know, it will be very surprising if they do not have the same general properties we have seen so pervasively so far: genealogy, divergence, duplication, reticulation, modularity, sequestration, interaction, functional use and reuse, and chance.

Early in their lives, Darwin and Wallace described the world as a self-directed phenomenon of change driven by competition, unfolding to the present panoply of complex organisms. Later in life, both men had their doubts. Darwin persisted in thinking of a Creator responsible for the initial start who perhaps then left things to go their own way. Wallace studied spiritualism and could never believe that the human mind could be the product of natural selection. Were these the bedside conversions of men facing death or the wishful thinking soft-headedness of age? Unlikely. It is clear from the work of both that these thoughts were present from the beginning. And their brilliant minds are not alone in this experience.

However it works, after billions of years in the making, evolution has produced such grandeur, so difficult to reconstruct in retrospect that it is difficult even for many evolutionary biologists today to accept completely mechanical explanations.

The most reductionist of molecular reductionists have sometimes struggled not to invoke teleological notions but to substitute what comes within a hair's breadth of doing so, for example, via a very determinative view of selection, or Jacques Monod's use of "teleonomy" and the "project" of life to explain a kind of inherent property, if not drive, in the very molecules of life (Monod 1971). René Descartes, who is often credited with starting reductionist science in the first place, said that even if an organism is a machine, it was driven by the spirit. Whatever that is.

Darwin suggested in the Descent of Man (Darwin 1871) that despite its importance he may have placed too much stress on the role of natural selection but that he did so to show that natural processes would suffice to explain the living world without the need for special divine intervention. In this book, we have pointed out ways in which selection may be somewhat less necessary than Darwin stressed because other natural processes contribute to plausible explanation. It is remarkable that so modest a number of basic principles can account for both the production and evolution of the diversity of life we see on Earth today.

One of the most consistent findings, and a continual source of doubt, is that there are exceptions to these principles. Indeed, viewing them as exceptions shows a danger in scientific inference. The exceptions can only be viewed that way if we take our rules too seriously, thinking of them as classic "laws of nature." That there will be exceptions probably is itself one of the fundamental laws of the nature of life.

0 0

Post a comment