Genetic
Predeterminism and Narrative Teleology
Presented at the MLA Conference, December 30, 2001
The concepts of species and self depend upon a notion of identity, a sameness that continues as a population changes or as an individual develops or ages. I'm going to be talking about whether or not identity is something that is predetermined by genetic make up. But I want to explain a little about my background first. I'm a narrative theorist, with an interest in causality. I think stories are basically models about cause and effect relationships. In 2001, I was a visiting researcher at the Santa Fe Institute (www.santafe.edu), one of the leading research centers for "complexity science." One of the things studied there is genetics. My research actually involved evolutionary mechanics. But, there is overlap in these areas in what is called spontaneous self-organization. You could say that the main interest at SFI is the question, Where does order come from? The short answer is that, in a way, it comes from disorder. In the course of my talk, I hope to make it clear how a stable identity and physical form can arise spontaneously out of largely stochastic processes involved in development. The so-called genetic program is not at all like a simple digital computer program that unfolds step by step in time. (Genes may, however, work like the new kinds of computer, e.g. quantum computing, but that's another paper.)
First some important terms need to be defined before I go on. Stochastic means random. Think of molecules in thermal motion. They bounce around in unpredictable ways because their movements are caused ultimately by quantum fluctuations. If these molecules bond to each other in different ways depending on whether or not, say, their left or right sides meet up, then their overall configuration will depend on chance. We would say, their organization would be the result of a stochastic process.
Spontaneous means intrinsically caused. Molecules in thermal motion form configurations all on their own without any help from an external molecule director or predetermined detailed plan. Spontaneous does not mean uncaused.
These terms can be used to describe other patterns besides those made by molecules in thermal motion, individuals trading in a stock market, for instance, or animals competing for resources in an environment. Although the individuals would not be in "thermal motion" their individual actions would be relatively random because not directly synchronized or organized by a central agency (such as Adam Smith's invisible hand).
Let me begin, now, by painting a picture of the way -- what I will simply call -- postmodernism conceives of identity. This is the first of two strawmen that I intend to demolish, by the way. They are radical indeterminacy and predeterminism. My own argument tends to the middle ground of what is called emergence.
According to postmodernism, to define an identity or assign a label to a human being, is to artificially fix a dynamical process and to ignore important differences. Identity is a state in flux that involves many individual elements that are not directly organized according to any known laws. The postmodern notion of identity enjoys the same status that Roland Barthes said narrative meaning has. In the spirit of Barthes, we may say that a self is a multi-dimensional space in which a variety of identities, none of them original, blend and clash. (See "Death of the Author"). These multiple identities might each be caused by one thing or another but the whole as such is contingent upon the accidents that make up a life. One sort of imagines these identities swarming and relating to each other only in a stochastic manner such that no one could ever predict or rationalize.
Nevertheless, the Human Genome Project launched in 1990 promised to reveal the blueprint that determines what and who we are. This was a rigidly predeterministic view. I think it's safe to say that most biologists aren't postmodernists. The Genome project was inspired by the belief started in the late 50s that what a gene does can be inferred from its structure. In the words of Francis Crick, (who won the Nobel in '62 for his work on DNA): "DNA makes RNA, RNA makes proteins, and proteins make us." (quoted in Evelyn Fox Keller, Century of the Gene, 54). In this view, the development of identity is linear and directly caused by the physical structure of DNA.
Perhaps some of you have seen pictures representing genetic transcription. A strand of DNA looks something like a spiral zipper, with teeth that are referred to by letters, A, C, G, and T. When it unzips, free-floating molecules attach to the teeth of, say the left, strand, creating a fair copy of right strand of the DNA. This copy is called RNA (RNA has a U where DNA has a T). Then bits of the RNA strand move from the nucleus into the cytoplasm of the cell where they function as templates for protein. Other free-floating molecules attach to RNA strand in sequences that define proteins. Of course, I'm simplifying, but this is the basic idea.
In this picture of DNA transcription into RNA and translation of RNA into protein, each part of the puzzle is exact. There is no room for error. If this is the way things work, then there is not much in arguing against genetic predeterminism. We are our genes.
Evelyn Fox Keller critiques this view in her book, Century of the Gene (Harvard, 2000). Keller's book forms the basis of my talk. She was trained as a theoretical physicist at Harvard, spent many years working as a mathematical biologist, and is now a Professor of Philosophy of Science at MIT. Keller points out that it would be a disaster if development did rely on such an inflexible model, for any tiny error in the beginning of the process would lead to exponentially greater errors later in the process. (It would be like Microsoft Windows, always crashing.) Development must have built in redundancies and ways for the process to correct itself through feedback. The order that arises cannot be uniquely determined; it must be probabilistic.
Since the earliest days of biology, researchers have wondered how the organizational plan of an animal (its identity,) is transmitted from parent to offspring. Then DNA was discovered and described. Here was an actual physical object that was capable of acting as a template. This seemed to explain a lot, if the object is handed off more or less intact to offspring.
Keller explains how in the 70s and 80s, the gene was described as a piece of autonomous matter with vitality and agency: Even to the day, the gene is sometimes referred to as the cell's brain. The fact that a gene can virtually reproduce itself during the process of transcription made the gene seem alive. The capacity to reproduce is traditionally taken as the defining property of life. Furthermore, we tend to think of genes as doing something not just being a code, since we have been told that genes make proteins. According to Keller,
endowing the gene with the power to act added to the property of materiality the further implication of agency. ... Attributing to the gene the capacity to direct and control development effectively credited it with a kind of mentality--the ability to plan and delegate.
In this view the gene was
Part physicist's atom and part Platonic soul. (47)
Keller claims that this view of a personified gene is no longer tenable. Furthermore, she claims that even the concept of a gene as a physical thing-in-itself has been "deconstructed" by recent studies. So in this sense at least, the postmodernist view of identity can be defended.
In her description, the code is transcribed in a process that is similar to the way the "fitness" of an animal is selected in Darwinian evolution. A lot of errors are made in the transcription process as well as in translation. Many of these errors are spontaneously self-corrected later on as the chemical environment changes and more of the necessary molecules are available.
At different stages of development, and in different parts of the organism, the "same" DNA can result in different proteins. Conversely, many different genotypes can produce one kind of phenotype. There is no simple relationship between gene structure and protein, and, what's probably even more important, there is no simple relationship between any given protein structure and its function. Both depend on context. Geneticists are starting to think that the gene just supplies raw material. And that it is actually the dynamical behavior of the materials that determines the form of an organism. (71)
Some interesting experiments have been done with self-organizing properties of chemicals that support this idea. In these experiments, genes are not needed to direct the organization of the pattern. A stochastic reaction-diffusion process in a petri dish can create the same kind of patterns that are found for example in animal fur, butterfly wings, and variegated leaves. Patterns form spontaneously if the chemicals are present in the right amounts. I can recommend two books on the topic: Phillip Ball's The Self-Made Tapestry and Brian Goodwin's How the Leopard Got its Spots. To give a very quick example, a totally simple model, let's say we have two chemicals, A and B. They can react together in one of three ways depending on which of their sides, left or right, come into contact. A and B can become two Bs. Or two As. Or remain A and B. Because the molecules are always in thermal motion , the way they happened to meet up is random. It is equally possible A and B will meet up such that the result will be A and A, B and B, or A and B.
One might think that together these reaction scenarios would tend to average out, maintaining the mixture in homogeneous steady state. This would be what one would expect according to the second law of thermodynamics. But this is not what happens. Instead differentiation occurs. If a small clump of As happens to form in one area, the As in the interior of this clump can only randomly interact with each other. No Bs will be produced here, since Bs are required to produce more Bs. The reaction continues as more As are produced at the edges. The clump grows. What one sees looking down at the petri dish are spot patterns, such as occur in animal fur. (Note that no chemical reactions are this simple! Think of this example of a cartoon illustration that conveys a general idea.)
So we see that, as postmodernists have argued, context and contingency is very important, even key. There is no central directing agent that directly causes patterns. Nevertheless, it is also true that some fairly predictable patterns do arise out of stochastic processes. In fact, nature is quite constrained in the kinds of forms it produces through spontaneity. These forms are known as structural archetypes or strange attractors. [1]
These days we attribute to the ordering tendencies of chance what before had been attributed to final causes, a divine architect or a detailed genetic code script. In teleology the question we ask about nature (is it goal-directed?) is the same question that we ask about humans when we are trying to decide whether or not we are intentional beings. I think the answer is YES, we are intentional beings, but we shouldn't think of goals as real targets in the distance that we aim for like arrows released by an archer. Instead, we should think of a leaf on a river being sucked into a little whirlpool. No matter where the leaf is launched from the shore, it almost always ends up in the same place. One word of caution, however, now that you have the image of the whirlpool as strange attractor, try not to think of the whirlpool as existing before the leaf gets to it, then you will have some idea of what this new kind of teleology might look like. (For more of my work on teleology, see http://www.dactyl.org/directors/vna/cv.html.)
What all this is saying is, that even in a world governed by chance, not all genetic narratives are equally possible. And if you will permit me (gracelessly) to extend the argument, regarding narrative meaning, not all interpretations of a text are equally possible. Some interpretations turn out remarkably like the parent text; others even improve upon it; but there will always be those that simply miscarry. You could say that, sometimes small differences in where the leaf is launched from the shore sends it off in an unpredictable direction, sometimes to a new attractor.
Order, pattern, structure need not have an a priori cause or a metaphysical presence in order to be considered in some sense objective. These ideas contradict Derrida's arguments, restated recently in Politics of Friendship, where he claims natural categories, concepts of identity, community, culture, and nation, are constructed by language. The problem with these linguistic conventions, he says, is that they cover up differences. According to the new genomics, the process of identity formation depends upon the idea that the individual elements are different. In order words, that they do behave stochastically.
Regarding the claim that identity (as a pattern) is constructed by language, an area of research known as computational mechanics addresses this concern. This area of study, which investigates the way nature computes or the way nature processes information, was developed by Jim Crutchfield (one of the heroes in James Gleick's book Chaos). Before computation mechanics an emergent pattern was recognized by an outside observer who imposed a preconceived model upon the pattern. Thus, you couldn't really say it was objective. This is as Derrida claims. The pattern is constructed by the observer's external model. Now physicists have folded the observer into the system, so to speak. Their "model" of the pattern is found in what is called the "causal architecture" of the process itself, that is, the procedure that produces the pattern. In this way, emergent patterns can be quantified and categorized with relative objectivity. (Unfortunately there are no accessible books or articles on computational mechanics as yet. But if you're brave you can look at http://www.santafe.edu/projects/CompMech/ .)
The more we learn about genetics, the more we can look forward to a holistic approach that argues that stability is an emergent phenomenon. As Keller notes, "This is a far cry from the old view of the gene as an inherently stable molecule subject only to occasional random mutations." The gene is dynamically stable. Likewise identity is dynamically stable.
In conclusion, I just want to say that the study of self-organizing systems -- be they genes, chemical reactions, or economic communities -- offers a new way of explaining the origins of order. One does not have to resort to Platonism, and one can allow for an idea of order that includes so-called deviations as part of what is normal and natural.
Victoria N. Alexander
Dactyl Foundation for the Arts and Humanities
64 Grand Street
New York, NY 10013
alexander@dactyl.org
www.dactyl.org/vna/cv.html
[1] Attractor is term, coined by Edward Lorenz, which, as a noun, implies a kind of thingness and imbues the concept with the sense of an actual independent force, and which I find to be every bit as misleading as any of St. Thomas Aquinas' terms for a telic principle.