Evolving Creative Minds - Decline of the Genomic Avatar

Charles J. Lumsden 23.04.1998

How could the emergence of the human mind be explained by evolution? - Part I

Does a passion to create drive the human species, making us utterly different from all other living things with which we share the planet? Or do our capacities for novelty, great and small, link smoothly to those in other species, so that human creativity is really a variant on a theme repeated countless times in the history of life? Are we to understand ourselves as expressing a "regional dialect" for innovation - unique and special in its own way to be sure, but nonetheless a restyling of universal evolutionary stratagems?

Charles Lumsden
Email
Homepage

Stressing descent with modification, Darwinism seems to say that we are both special and mundane. Is this a fact of human evolution or a fact of evolution's inability to explain humanity? Partially humbled by our place in a teeming pattern of life on a small planet lost in the stellar haze of a nice, but typical, spiral galaxy, Darwinism nevertheless seems to reach, Prometheus-like, past psychology and even past philosophy to explain our behaviors, minds and social forms in a language once reserved for debates about hybrid corn or fungus growing among ants. What is going on?

Anzeige

Figure 1. The cognitive revolution: Can we find ourselves in the hardware? Drawing by Nick Woolridge.

Blame it on the cognitive revolution. In the breezy days of behaviorist populism, evolution, following Darwin's lead, was content largely with speculations about the origins of instinct and drives. This was the thin of the wedge evolutionists have driven into the mind as behaviorism has thawed into cognitive science and the mysteries of consciousness, intention, self-awareness and human genius returned from the fringes of scientific respectability to crowd the center of human science. The linchpin, of course, has been the aggressive detente set up between the brain scientists and the mind scientists. The wall is down.

Increasingly, claims to a proper study of humankind are taken as the common, if controversial, property of both parties (Churchland 1995 is good example - one of many recently - of epistemic detente in action). As we will see, biological evolution withers without organic matter, filled with heritable variations, on which to work its seeming magic. But once the mind-brain connection gains a toehold, evolutionary exegesis follows swiftly, and has, buoyed by a remarkable volume of anthropological, ethological, paelontologic and genetic fieldwork among human and animal populations.

What is creativity? Resembling evolved organisms to an uncanny degree, characterizations of creativity bear the unique imprint of their progenitors along with a mild degree of family rsemblance: creativity as a kind of capacity to think up something new that people find significant (e.g. Amabile, 1983; Boden, 1991, 1994; Sternberg, 1988; Sternberg & Davidson, 1995). It is possible to be more precise than this, of course, and in order to maintain consistency with my earlier usage (starting with Findlay & Lumsden, 1988) I shall by creative process have in mind those mental events by which an organism intentionally (Dennett, 1996) goes beyond its prior experience to a novel and appropriate outcome.

Creativity will refer to that tantalizing constellation of personality and intellectual traits shown by people who, when given a measure of free rein, spend significant amounts of time engaged in the creative process. Outcomes achievable in principle by creative organisms can vary hugely in their novelty and significance. The Wright brothers, for instance, could have stayed home and made better bicycles instead of undertaking their momentous journeys to Kitty Hawk (Bradshaw, 1996; Freedman, 1991), with all the difficulties and deprivations those entailed.

An outcome is a product of the creative process. Bach fugues and Gödel numbers and our kids' WEB pages are all outcomes of the creative process in this sense. The creative process need not have an outcome consistent with the organism's originating intentions. But it can, as the invented world that is home to our species demonstrates. In science, outcomes often teach us something about the world that predates our intentions and are called discoveries. "Discovery" seems less apt a moniker for outcomes like paintings and rock concerts; "works" is more common.

An innovation, finally, refers to an outcome that attains some level of adoption in the society under consideration. Outcomes can fail to turn into innovations, or be lethargic in making the transition through a lack of attributed value (so only lately has van Gogh found fiscally astronomic value in the eyes of the marketplace) or because the discoverer "sits" on the breakthrough, or through sheer happenstance ("market share;" remember VHS versus Beta?). Can evolutionary science offer any insights into what is happening in such diverse circumstances, over and above that already possible through psychology and the neurosciences?

I think the answer is a guarded "yes" because evolutionary science treats questions of special relevance to creativity research. These are the "why" questions. As in the behavioral sciences, much of biology is concerned with "how" questions: how cells divide, how long term memories are stored in the brain, how parents socialize offspring. These are familiar questions dealing with process and mechanism. Evolutionists, however, want answers to the "whys" rather than the "hows" of anatomy, physiology, and behavior: why cells divide in a certain way or why parents sometimes behave altruistically toward their offspring.

Figure 2. Creativity wetware. In the human brain, here seen from below, a richly folded neocortex mixes sensations, memories, and ideas into the self's desires and goals. How? Why? Painting by Nick Woolridge.

The question "why?" demands the study of history - the history of the individual (development, socialization, learning, choice), their culture and society, and ultimately their population. The history of biological process (the thin end of the wedge, the biology of brain driven by evolutionists into the mind) is by definition evolution. Let us consider this latter point more carefully, because above all it is important to know why human creativity is as it is, rather than some other way it might have been (why aren't we all Blakes or Bachs?) had the past been different.

Splice and Dice

The modern or neoDarwinian theory of evolution (NDT hereafter) is a synthesis of Darwin's natural selection (Darwin, 1859) with Mendelian genetics and the population biology. Codified in a series of masterworks by Dobzhanksy (1937), Fisher (1930), Haldane (1932), Mayr (1970), and Wright (1968), NDT by 1950 rested on a broad base of empirical and mathematical discoveries that continues to be deepened and elaborated. Although the exegesis on NDT is endless (exciting surveys with some emphasis on behavior are Brandon & Burian, 1984; Dawkins, 1986; Dennett, 1995; Maynard Smith,1982; Williams, 1966, 1992; D. S. Wilson, 1980; E. O. Wilson 1975), its core idea can be immediately grasped and put to work here.

For neoDarwinists, evolution is a dramatically creative, albeit non-intentional natural process pivoted on a tension between the genesis and the shaping of raw diversity. Even though the chemical reasons were not grasped until the advent of molecular genetics in this century, it was well known in Darwin's time that many traits (such as the color of your eyes or your ability to roll your tongue into a tube, to cite current examples) passed in a seemingly deterministic fashion between parents and their offspring. Individuals differed from one another, and some of these differences were heritable.

In retrospect and filtered through 140 years of refinement and exegesis, Darwin's key insight today seems the model of obvious simplicity: if a heritable trait varies among individuals, and if a change from one version of that trait to another also changes (increases or decreases) the organism's reproductive ability, then over time we can expect to see those versions of the trait associated with greater reproductive ability become more common in the population. This is because the individuals with these versions will have the greater number of offspring, which inherit the version and in their turn leave the greater number of (grand)offspring, and so on. If, moreover, individuals must compete for scarce resources, such as nourishment or shelter or mates, then over time many of the other versions may vanish altogether. This is natural selection in action.

For example, suppose that in a hypothetical species individuals vary in their ability to taste a particular poison, owing to differences in genes at one or more places on the chromosomes. The gene codes, say, for a cell-surface receptor protein able to bind molecules of the poison and trigger neuronal signaling, conferring the ability to taste the poison. The opposing gene variant provides no such ability because the receptor site of the protein it encodes does not bind the poison at all. The poison is present in the food or water. Individuals with the "taster" variant of the gene detect the substance, avoid it, and survive. Those possessing "nontaster" versions consume the poison and die. As a result, the frequency of the taster variant increases in the population, and a larger percentage of the population has the hereditary ability to avoid the poison. This change from one generation to the next is evolution by natural selection. In the vernacular of population biology, the individuals able to taste the poison are said to have "increased genetic fitness" and the taster trait is spoken of as "adaptive" relative to the population's specific circumstances.

Many factors can change or disrupt the otherwise systematic shift over time in the relative abundance of these versions. Environmental change can reverse the reproductive impact of specific traits. Traits adaptive in themselves may interlock in unexpectedly deleterious configurations. For example, the "taster" gene variant might also confer a drastically reduced sensitivity to sexual pheromones, so that even though fewer tasters die from the poison than do nontasters they reproduce far less, so that over time it is the nontaster rather than the taster version that increases in frequency. And so on.

Natural selection culls diversity. No heritable diversity, no natural selection. But where does the organic diversity, the evolutionary raw material, come from? At least four processes are known to sustain the continued appearance of new gene variants or new patterns of gene organization in biological populations:

	mutation, physical events that change one or more letters of an organism's nucleotide text;
	recombination, physical events that mix and match gene variants among parental chromosomes during sex cell formation;
	migration, the arrival in a population of individuals carrying novel variants of genes; and
	sex itself.

Figure 3. Paths to genomic diversity. Mutation rewrites the DNA text while recombination cuts and pastes DNA fragments from one genome to another among breeding organisms, eventualy producing multiple versions of an original gene. In this highly schematic illustration, a gene labelled 0 eventually gets diced and spliced into variants 1 to 9, each with a relative fitness value (shown abstractly by the color). Illustration by Charles Lumsden.

For NDT the primary role of sex is more subtle than straightforward reproduction or behavioral frolic. It is the creation of genetic diversity among offspring. An organism that reproduces without sex, say by hatching unfertilized eggs, can replicate its genome exactly, gene by gene, without wasting time on the rituals of courtship. But if all offspring are identical (mutation helps asexual organisms assure that they are not, but pales compared with the creative powers of sex), they may be less likely to withstand important changes in the environment. Sex mixes the genes of at least two sometimes consenting participants and endows offspring with more than one copy of each gene, which may differ from each other and therefore hedge against hard times. Sex can be slower than nonsex in terms of the rate at which natural selection can change gene frequencies within the population, but it provides an more diverse array of genetic combinations to present to the world.

Sex also underlies recombination, a chemical dance among parental chromosomes in which whole chunks of DNA from one parent get swapped with those from the other as the chromosomes are sorted, lined up, copied, and finally shunted into place in the sex cells. A very potent recombinant event is gene duplication, where two or more copies of the same gene get inserted into a chromosome in this molecular version of cut-and-paste. Freed from their normal role in cell physiology, the backup copies of the gene may diverge from their ancestral structure with fewer or no immediate harmful consequences for the organism, allowing new gene products to form and enter the cellular matrix. Many of the proteins that control cell activity in your body are thought to trace their origins to duplication events among genes or parts of genes.

The masters of NDT knew mutation and recombination as diversity-creating processes indifferent to natural selection: the physics by which genes splice or one nucleotide base becomes another (or is mistaken for another during gene replication) goes forward unruffled by the adaptive struggles of the organism. Of course mutation and recombination are not necessarily arbitrary in the sense that any conceivable variation in an organism's structure and function can be produced by just one mutative or recombinative step. The biochemical control net of development may damp out a change or shunt its effects toward specific outcomes. NDT requires only that the effects be arbitrary relative to the direction in which natural selection is acting. Then, over time, the cumulative change in the population as more and more mutative and recombinant events occur and are passed into succeeding generations might be very great indeed. (Sex, as usual, is even more interesting; what sex cells get together is very much tied to natural selection and has given rise to sexual selection, a major evolutionary subject in its own right starting with Darwin, 1871.)

Would it not be handy then, for an organism to get physics working more directly to its adaptive needs, in the sense that it could more often mutate those genes whose change would improve the relevant traits? Why waste time changing eye color when there is a poison to be detected or successes to be had by being more creative? Such closet Lamarkism is anathema to NDT. In the autumn of 1988 John Cairns, Julie Overbaugh and Stephan Miller rocked the neoDarwininan apple cart when they reported, in Nature, experiments on the humble bacterium Escherichia coli purporting to show mutation influenced by adaptive "need." If E. coli with an inoperative version of the gene segment Lac, the working version of which gives the cell an ability to digest the sugar lactose, were put in a broth containing lactose as their only carbon source, the cells seemed to preferentially crank up Lac's rate of mutation. As Cairns et al. (1988) interpreted the situation, the bacteria detected the presence of lactose and channeled their dwindling resources into producing mutations most likely to help them out of their adaptive jam.

The response to the work of Cairns et al., along with later experiments alleging evidence for directed mutation (also for the most part in bacteria) has been swift, sharp, and on the whole decidedly negative, focusing on potential flaws in the experimental designs as well as alternative, traditional interpretations for the data. Cochrane (1996), Keller (1992), and Sniegowski and Lenski (1995) survey this lively topic. Big epistemic stakes obviously ride on directed mutation, and it is clear that the null hypothesis of random mutation will be stoutly defended. For the moment, the nature and evolutionary prominence of "directed" or "adaptive" mutation is, despite its fascination, best regarded as highly provisional and little understood.

Will That Be Micro or Macro?

Darwin opined that environments could be sufficiently stable to allow natural selection to operate over long times, producing gradual but cumulative differences among biological populations. The creation of small amounts of difference by natural selection over relatively short periods of time is no a longer terribly controversial issue. This is microevolution.

Microevolution sometimes proceeds by means other than natural selection. Mutations can occur at such a high frequency as to elevate the percentage of mutants in the population without the effect of natural selection. Immigrants can bring new genes into the population at a high enough rate to change its overall genetic composition. In small sexual populations, the statistics of assortment among recombinant genomes can itself become an important determinant of the relative abundances of the gene variants. These phenomena occur and are at times significant, but current evidence indicates that they are much less potent than natural selection in directing evolution over longer periods of time. In other words, natural selection is the dominant mode of directed change in microevolution.

Natural selection ultimately must work on gene changes trickling through the cell biology of organ development, and through any limitations the physical laws of chemical reaction place on the form and interactions of cells, tissues and organs (e.g. Goodwin, 1994; Kauffman, 1993). The nonlinear properties of the biochemistry and the structural physics of the materials and forces holding an organism together are of special importance in this regard. In nonlinear systems, what you get out is not necessarily proportional to what you put in; it can be wildly different, organized into specific ranges of exotic pattern and behavior (an excellent introduction to nonlinearity is Kaplan & Glass, 1995). Thus otherwise small changes in genes, acting through the nonlinear "constraints" of developmental biophysics (Goodwin, 1994) and molecular biology (Raff, 1996), might be of large effect in organismic development, or be damped out by the nonlinear cushioning even more than we might expect.

Figure 4. Macroevolution. More than 10 million years of cumulative change separate the compact brains and minds of modern apes and monkeys from the imaginative consciousness housed in the human skull, and from its legacy of material artifacts. Skull drawing by Nick Woolridge; Charles Lumsden photo.

Far more controversial than microevolution is NDT's capacity to explain macroevolution and the origin of species: its proposal that, properly extended, continued gradual accumulation of small differences under natural selection ultimately could create new species (populations of organisms isolated reproductively from each other), and beyond that the higher-order taxonomic clusters of body plans and behaviors recognized to order the diversity of all living things. The evolution of creativity has strong macroevolutionary overtones. Play and innovation behaviors are not unique to us but occur in many species, especially those with larger brains, warm blood and complex social structures (Fagen, 1981 is the classic synthesis). The evolution of humankind also is a macroevolutionary process in itself, a succession of bipedal species within the genus Homo that fades gradually into the bones of long-vanished primate ancestors.

The wrangles over macroevolution have been heard as announcing the death of Darwinism. NDT, like history and God, is allegedly at an end (e.g. Goodwin, 1995; King, 1996). Not so. Modern work has placed NDT, and with it gradualism and the concept of natural selection, alongside alternative hypotheses about the large-scale evolutionary mechanisms at work behind the fossil record and ecological diversity. It would be premature to claim that data currently are sufficient to allow any of these hypotheses to be rigorously posed, tested and refuted. There are hints of a whole spectrum of temporal patterns across which evolutionary change may be arrayed, from smooth gradual progressions in some cases to intricate patterns interweaving long periods of little or no change (so-called stasis) with shorter bursts of evolutionary innovation (Raff, 1996). The play of developmental constraints against environmental shifts and genetic variation may figure in the transitions between stasis and change in such modes of "punctuated equilibrium" (Gould & Eldredge, 1977; Eldredge, 1989). As we will see below, this includes our own fossil record of human creativity.

As the debates over macroevolution continue, it is important to keep in mind the difference between change per se and directed change genomically ad hoc leaps into the unknown versus the systematic remodeling of populations over time. Mutative and recombinative mechanisms, for example, excel in ad hoc leaps, possibly scrambling the biochemistry of organic development to produce really novel forms. They may figure in circumscribed intervals in the history of life linked to sudden bursts of diversity.

Apart from the barely understood mechanism of directed mutation, the usual suspect for systematic change is natural selection. The integrative nature of development makes it possible, however, that some trait we deem adaptively important actually is going along for the ride because ontogeny constrains it to change when the trait, on which natural selection really is acting, changes. Evolutionary biologists sometimes call these free riders preadaptations, and they are interesting because they can reshape evolution by suddenly opening adaptive opportunities uncorrelated with previous adaptive trends.

Any Genes With That?

Just as Darwin worked without a proper theory of inheritance, the principal synthesizers of NDT and its subsequent applications to behavior and population dynamics (Dawkins, 1976; Maynard Smith, 1982; Williams, 1962; D. S. Wilson 1980; Wilson, 1975; Wyne-Edwards, 1962) were forced to work without a proper theory of organismic development. The link between gene activity and the "finished" organism had to be fudged though an almost hieroglyphic shorthand, a statistical algebra that skipped over ontogeny.

The direct link, if there is one in a mathematical sense, is between a change in a gene and a change in the organism (and thus in its pattern of development). There is no such link between a gene and a trait per se: genes code for gene products (other molecules that regulate the genome, or proteins active in the structure and physiology of the cell). Nothing else. The unfolding of the organism in development passes the information of the genes into the matrix of interaction with the cellular material of the preexisting egg, where the conditions of the environment and laws of chemical physics enter the picture, governing the flow of material and chemical reactions. The molecular language that genes speak to each other is now partially understood by developmental biologists, allowing them to trace with increasing precision the action of individual gene products in shaping the organism.

Thus, there are no genes "for" sexual preference or enjoyment of clean environments or musical precocity, although there may be genes in which a change in turn changes such traits, perhaps significantly. The information-rich stretches of nucleotide text in DNA translate into ribonucleic acid (RNA) which may in turn be read off into strings of amino acids that fold to make proteins in the cell, or into sites through which the transcriptive activity of yet other DNA locations in the genome can be regulated (turned off or on, or increased or decreased in the their likelihood of entering an active or in active state). Regulator sites on the genome can regulate yet other regulators and so on, making complex chains of control and a strong capacity for self-organizing activity of the genome in its interactions with the rest of the cell (Kauffman, 1993), and within cell populations during development.

It is tempting to simplify such a picture of distributed nonlinear genetic interaction and integrated genomic activity by a conceptual elision that makes the organism and its expressions (including the human mind) look like avatars for their genomes, stiriving endlessly to enhance reproductive success. But unless contrued as a rhetorical flourish, the elision of "genetic change --> trait change" to "gene --> trait," is incorrect and rashly misleading. Talk about genes of "major effect" means that within a target population differences in a trait among individuals can be traced in significant part to differences among their genomes, not to differences in the environment (or in the laws of physics sustaining their ontogeny and physiology). Such a claim is not equivalent to asserting that the specific trait is determined or blueprinted in the gene itself, or that the environment or chemical physics of development is irrelevant.

The discourse of molecular biology at times equates the information content of a genome to a sort of blueprint (Goodwin 1994, 1995; King 1996), a polynucleotide ribbon holding little snapshots of the organism and its parts just like your undeveloped film holds still-invisible images when you drop it off for processing. This seems to prepare tot field of battle for imputations of genetic reductoinisn of the most vulgar, albeit full-blooded, sort: To build an organism all we need, apparently, is the DNA sequence and the right "soup". But even in fast-paced fantasies like Jurassic Park, you'll recall that the carnosaurs did not spring phoenix-like out of heaps of coiled DNA: the nucleic acid had to go back into the chemical soup of the embryonic environment, joining with other biochemicals to form a cell that grew into hungry predators and big profits. The DNA is not a blueprint except in the clubhouse stretch of a tired metaphor, but it is a principal subsystem of development's tempos and modes, the one to which we look in most organisms to explain the origins of heritable variation. Changes in genes can change ontogenetic tempos and modes, and with them physiology and behavior.

Short Circuits

Seen through the peephole of statistical correlation across generations, heritable variation among humans has another source: culture. We live in a world of invented meaning, a devised world our species has created for itself, surrounded by social things of our own making. Crisscrossing the lines of genetic inheritance by which humans reproduce their bodies is the flux of culture; deprived of it we have human bodies but not human minds. This our niche. Whereas most animal evolution arises from the differential replication of genetic information, human evolution is tied up with the differential transmission of both genetic and cultural information. I readily admit that it does culture no full justice to force it into partnership with a desiccated term like "information." I shall do so here only because I see no other that works quite as well within the space constraints at hand.

Cultural information refers very loosely to the stories, ideas, and behavioral stratagems shared among people in every society on the planet - and without which we are unable to be in the world in a manner that is meaningful to each other, and perhaps to ourselves (Randall, 1995). Jumping into a laguage game with a term like "information" in our TELEPOLIAN age makes it sound as though culture is particulate in the way our genome might be, that twisted ribbon of discrete DNA base pairs, ultimately is. But culture, and the action of human creativity on it, are replete with information that is continuous and irreducible, whose potential meaning and significance cannot be exhausted - paintings, texts, symphonies, and so on. This gives culture a density of packed meaning unknown in the genetic world (Lumsden, 1989). Changes in the abundance of a variant of a story or invention over time is cultural evolution is it simplest, currently tractable form.

Anthropologists and evolutionists rightlfully demand to know how independent genetic biological evolution and cultural evolution have been (and are) as they have run forward from out primate beginnings 5-7 million years ago. My study of data drawn from developmental psychology, cognitive science, and the comparative ethology of animal societies that show some rudiments of cultural transmission (Bonner, 1980), offers tentative support for the hypothesis that human culture and the human genome are not evolving independently on their own, isolated tracks. The neurobiology of culture learning makes them codependent, resulting in the process of gene-culture coevolution (Lumsden & Wilson, 1981).

Gene-culture coevolution in human beings appears to be organized around a mechanism of information inheritance termed gene-culture transmission. In gene-culture transmission, genome activity picks a special mode of organismic development. The neuronal learning rules of gene-culture transmission influence the likelihood of some rather than other variants of culturalinformation causing large changes in the child's pattern of enculturation. The reductive nature of the term "information" makes it sound as though this statement deals with patently absurd alternatives such a preference for blue jeans rather than black jeans being carried in the learning rules, but this is not the meaning at all: the core issue is the species-specific nature of our minds: our culture-dependent self-awareness, the grammatical foundations of our uniquely human language, our cross-cultural bias for brother-sister incest avoidance in courtship and mating ceremonies - systems of meaning and understanding acquired by children quickly, effectively and without intensive teaching.

Figure 5. Cultural evolution. Creativity assures that culture change during gene-culture coevolution is more than just a branching tree of diversifying ideas, lifeways, and artifacts. Key innovations, appearing unexpectedly, might significantly change a culture, causing it to "jump" from one line of cultural evolution to another. Aquatint by Dominic Hay.

In gene-culture coevolution a circuit of reciprocity operates in which genome subsystems affecting neurogenesis and culture learning shift in response to culture changes that effect the differential transmission of the underlying gene variants. While the analysis of mathematical models of gene-culture coevolution indicates that natural selection can strongly affect the rate and direction of coevolution, the results of natural selection can be markedly different from those expected on the basis of genetic reasoning alone. Developmental processes for culture learning, for example, can create nonlinear couplings between genomic and cultural change, with surprising results (e.g. Findlay, 1991; Findlay & Lumsden, 1988; Lumsden, 1984, 1985; Lumsden & Wilson, 1981). The diversity of evolutionary outcomes can increase, as can the rates at which they are approached. Altruistic and cooperative behavior can spread through a population without the aid of kin selection, reciprocal altruism, or any of the mechanisms traditionally brought to bear on the evolution of human social behavior. Higher-order evolutionary processes such as group selection can be more important than if evolution is purely by genetic means.

Human creativity is the fire that drives gene-culture coevolution. From creativity flow innovations, the raw material of cultural diversity. Cultural evolution, whether considered alone or in the context of gene-culture coevolution, stalls flat if diversity is zero. The action of gene-coevolution on this created diversity, starting with culture's modest beginnings in pre-hominid times, has made us human. The physical record of gene-culture coevolution's track through time, the fossils and the artifacts, offers surprising insights into our species' quirky, obsessive penchant for innovating with culture.

References

Acknowledgements