Date: Sat, 8 Aug. 1998 (Summary Date)
Tuesday, July 21, 1998. 7 P.M. - 9 P.M. (Seminar Date)
John Fentress, Ph. D.
Department of Psychology and Neuroscience
Halifax, Nova Scotia, CA B3H 4J1
15 attended John's presentation and discussion in the family room at the speakers' house; a supplement to the course "Healing the Moral Animal: Lessons from Evolution."
Instruction versus selection models are now hotly explored in psychology, neuroscience and a variety of other biological fields. Nearly all workers today ADOPT a broad epigenetic approach. THIS APPROACH PLACES EQUAL STRESS UPON GENETIC AND EXPERIENTIAL EVENTS, AND THUS incorporates developmental conversations between defined intrinsic and extrinsic factors. IT is much more productive (and reproductive) THAN ARTIFICIAL DICHOTOMIES THAT ATTEMPT TO TREAT EITHER GENES OR EXPERIENCES IN ISOLATION. Adaptive behavior reflects the match between organism and environment at every phase of developmental trajectories, the outcomes of that match heavily influencing reproductive success.
Bob Wright said earlier, "One thing genes do is set up flexible behavioral programs." Mechanisms are unknown but may have been reflected in archetypes, ethology, or developmental neuroscience. Constraints are likewise unknown. There are roles and limits for experience. The challenge is to examine both developmental constraints and rules for plasticity within a unified context.
As a starter, it is useful to think of experience as a much broader concept than learning. Experiences are essentially everything impinging upon (and processed by?) the developing organism. Developmental boundaries open and close. Different genetic backgrounds can produce very different consequences from the same experience. Experiences may add information or select (and amplify) preexisting potentials. On the other hand, biological systems do not just "mirror" the details of experience, but amplify, distort, and elaborate upon these details. And there are times when organisms seem quite insensitive to events that have a major impact at others.
Instruction, selection, and support interact to produce "interactive, self-organizing systems" (ISO) that result in ontogenetic adaptations. ISO systems represent an attempt to bring together the two basic facts of organisms interacting with (and responding to) their environments, and also utilizing inputs from these environments to enhance as well as modify ongoing developmental processes. That is, developmental experiences should not be viewed as writing instructions upon a blank slate, but as modulating developmental processes that reflect both the organism's genetic makeup and its previous developmental history.
THE THREE LITTLE P'S:
I reflect my wolf bias through analogy to the story of the three little pigs. First, it is essential to deal with behavior as a process, that is to evaluate its dynamic properties over a number of time frames. Second, these processes are in some sense patterned (and thus representing both separable and interconnected events) rather than continuous (diffuse). Third, behavior is a phenotype. One can summarize this analogy by saying process is the straw house of behavior, with pattern the stick house, and phenotype the brick house. The idea is that different properties of behavior range from ephemeral and continuous events, through dynamic systems and finally stabilizing constraints. We do not understand well how these different aspects of behavior operate together.
There are immediate and remote influences on the outcome of individual development. That is, organisms are sensitive to selected aspects of their environment ("Immediate Environment) while being insensitive to other physical events ("remote environment"). Thus, it is important to detail which aspects of the environment are respondED to throughout development. The direction of response can then be documented. Even though the organism does not (by definition) respond directly to "remote" environmental features, these features can influence the "immediate" environment, and thus indirectly affect behavior, and its development. It is therefore often difficult to know where to look for potentially important developmental events.
DYNAMIC NETWORK ANALYSES (DNA)
DNA is of course the bottom rock in evolutionary psychology. Genes set the ultimate potentialities and constraints in adaptive behavior. While acknowledging this, the emphasis of the present seminar was upon dynamic networks that link experiences across multiple channels., including those that have strong genetic constraints. The gap between the DNA of genes and behavior is huge, and thus Dynamic Network Analyses are needed to trace the various pathways during ontogeny that make adaptive behavior possible. Such studies can be carried out at many levels, ranging from biochemical pathways through cognitive and social operations of great complexity. The search for generalizations that cross these levels of complexity, as well as the transformations that occur when the levels are crossed, is at a very early stage. Evolutionary psychology can offer important insights into which search procedures are likely to be productive.
The linear model of a gene "causing" a particular piece of behavior is misleading, despite use of the model in conversation and in the press. Dynamic Network Analysis implies that several genes may influence a common outcome; as such, the pattern of influence is often more stable than any single gene's activity. Likewise, each gene may participate in several different outcomes, each outcome associated with a different cluster of genes.
Tinbergen's concept of a tight relationship between evolution, individual development, immediate and distal causes for a behavior, and the function of a behavior is still useful. It is possible to apply this concept at varied levels (social/ecological, organismic, and mechanistic) and at varied times in the life of a creature or a species. Evolutionary history, thus, is expressed in different environmental contexts.
The issues in developmental (epigenetic) analysis can get complex. Within this complexity is the basic fact that experience and genes can interact in multiple ways. If one starts with the very beginnings of differentiation in development, it is clear that experience and genes are in some sense inseparably linked.
For example, the cells in our noses, eyes, and ears are for all practical purposes genetically indistinguishable. Yet these structures, and their functions, are certainly distinctive! How does this happen? It happens in part through different experiences that precursor cells (cells that later become noses, eyes, and ears) have as development unfolds. These differences in experience come from many different sources, including the relative positioning of cells in the embryo. The cells have different neighbors, and thus different micro-environments. Different genes are activated in different patterns, and different structures/functions emerge.
A natural history example involves locusts. Locusts of certain species can occur in both solitary and migratory phases. Animals in these two phases look and behave very differently from one another yet they are genetically indistinguishable. How can that happen? In this case one key issue is nutritional state. If the animals have a nutritionally rich (complete) environment, they develop into solitary animals with heavy bodies and short wings. However, if early nutrition is compromised (such as through diminished plant material), then the animals develop into the migratory phase with slender bodies and long wings. What happens is that different genes are activated in different contexts due to differences in internal state. A given animal can become either a solitary or migratory locust depending upon experience. The transformation of potential worker bees to queen bees represents a conceptually similar example.
It is important to take both experience and genetics into account. Thus, Skomer voles which are notoriously tame are only tame if they have the correct genetic makeup and also correct early experience. Bank voles are genetically similar, but very timid. If a Skomer vole is cross-fostered to a Bank vole, it becomes skittish, as is a Bank vole either raised by its mother or cross-fostered to a Skomer vole. The only way one gets a tame vole is to have both Skomer genes and Skomer early experience.
Different subspecies of peromyscus (field mice) live in different habitats. Again, habitat preference depends upon complex (but rule given) combinations of early experience and genetics. The same can be seen for essentially all forms of adaptive behavior. Of course, some action potentials seem to be more tightly constrained than are others. Others are more flexible. Even here the issue is complicated by the fact that relative constraint versus flexibility can depend upon which particular variables are chosen for investigation.
The bottom line is that every action that an organism shows reflects its genetic potential. However, this is very different from concluding that experience is unimportant in promoting or dampening this potential. Further, organisms have multiple potentials, and strong expressions of some can block the expressions of others (as in the locust example).
As a pragmatic issue, science can only study differences between events. Thus, if one has two strains of mice, one might fight and the other might not. This difference can be taken to reflect a difference in genotype. However, this does not mean that fighting is genetically predetermined in one strain, or genetically precluded in the other. If either strain is reared under different conditions it might well be possible to produce one group of mice that fights and another does not fight. Obviously, one would not want to conclude that genes are irrelevant. It's just that experience is also relevant. The situation can get more complex, for which strain fights (or does not fight) can depend upon developmental events; that is, a fighting strain under one set of conditions might become docile under another set of conditions, and vice versa.
One simple conclusion is that organisms do not simply "mirror" their experiences, as if these experiences operate in a vacuum. The extent to which differences affect behavior, and the direction of that effect, depends on the organism's current phenotypic state, and that state in turn includes its genetic makeup and previous history of experiences. Genes do not dictate. They set up the conversations with an organism's world that allow the organisms to march through development, and perform adaptive (or maladaptively) at each ontogenetic stage.
Jim Brody, Ph. D.
I'm grateful for John's style and information, and of the path he's traveled to become what he is at this point in his life. The class and I appreciated his (and Mary Ryan's) giving so much personal time to the seminar and to the course, "Healing the Moral Animal: Lessons from Evolution."
First, I agree with everything John says.
Second, the nature-nurture acrimony, because of the plausible and sensible information that John provides and which is shared by an incredibly large audience, becomes interesting in itself. There is a group of us, whether by "inherited adaptations" or other mechanisms, who are more fascinated with biological contributions to human behavior than with the sticky webs of social explanations. Thomas Kuhn (Structure of Scientific Revolutions) remarked well that established science ignores inconvenient, discrepant observations. Lynn Margulies (and Max Planck well before her) commented that things change in science when some people (scientists?) die off and differently behaving people (scientists) take their place. Harsh language but they may be partially correct.
The polarity exists even though most of us are "interactionists" at our core; it may be simply that some of us "like" doing the nurture thing (e.g., Leon Kamin) and others "like" chemical tools (e.g., Sandra Scarr). It's difficult to escape primate thinking whether we analyze our behavior or whether we choose leaders. There's something about us that just "has to take sides." The level of emotionality that sometimes attends these discussions suggests motives related to dominance, hierarchy, and territory comparable to Superbowl fans rather than attaining an understanding of nature. (I can't recall who remarked that Nobelists are less characterized by seeking understanding and more for proving that they are right. Tom Huxley and Wilberforce, Bishop of Oxford, argued a lot with each other. Neither would have amounted to much without the other.)
Third, powerful effects are shown to be functions of environment interacting with a common genetic foundation. (For example, alligator gender is at least a partial function of egg temperature which is manipulated by mom's arranging leaves on her nest. The heat from decaying vegetation can make one egg a guy; absence of the heat tips the balance towards a girl from that same egg). At one extreme, we have great scientists showing "See what I can do to one egg by changing the mother, the diet, or the temperature." At the other side, we have equally talented people showing constancies in outcome with human beings despite environmental variation.
There are many possibilities. Many human traits COULD rest on input from a wider array of 3-10 or more genes rather than depending on just one. Multiple inputs should, in theory, make the trait more resilient, less variable than if it depended on one or two genes. Thus, the possibility exists that SOME human features could be more durable to variations in individual genes or to variations in environmental tuning than the traits shown by mice or voles.
Nonetheless, it's interesting to have the mix -- people working with smaller critters demonstrating the power of environment; people working with humans demonstrating the importance of genes -- which is a change in the traditional theory flow of "environment for people; genes for flies." In some senses, our morality could be as genetic as our grammar; but its expression remain heavily dependent on environmental factors. What fun!
Fourth: Statistical methods emphasize the detection of a difference; our minds often class events by similarity even without the statistics. A collector's case of butterflies, of shells, or of psychiatric diagnoses demonstrate this interaction between "same" and different." We put things in a different pile if they differ from things in our first pile; however, the events in the first pile are judged as "same." Pearson and Fisher gave us mathematical tools that emulate our sensory mechanisms for spotting differences; we need similar tools besides our psychological adaptations for reaching agreement on similarity.