The place of genes
Dear Lewis,
You and I have fenced over one of the central questions in biology-and indeed life-for many years now. It may be instructive to try to set out our differences more systematically. The question is how best to understand and explain living processes, in particular the role of genes within these processes. You, in common with many molecular biologists, are content to place DNA and its replication at life's centre. My book, Lifelines, is an attempt to rebut the claims of this genetic imperialism which, I contend, derives from mistaken metaphysics, misunderstands the nature of living processes and encourages potentially pernicious claims about genes for everything from violence on the streets to sexual orientation, drug addiction and compulsive shopping. But let's leave outcomes to one side for a moment and confront two issues: the limits of this reductionism and the meaning of genes.
Reductionism is a portmanteau word, but its thrust is summed up by James Watson, co-discoverer of the DNA structure: "There is only one science: physics; everything else is social work." That is, the task of science is to explain complex properties-such as cells, minds or societies-in terms of their component elements, which will ultimately render unnecessary any "higher level" explanations. By contrast, I argue that there is a hierarchy of organisation of matter which grows increasingly complex; at each level, new principles of organisation emerge which are not reducible to "lower" levels. Thus, temperature is a property of a collection of molecules, not a single one; living cells possess a structural order and stability which can be explained not in terms of component molecules alone, but in terms of organising relations between them; genes cannot be "selfish"; memory is a property of the interacting network of nerve cells which compose brains/minds, not of individual nerve cells; violence is a social process, not the property of an individual within a society.
Reductionism converts dynamic processes into static properties and structure into composition. Its methods may describe a living system, but they cannot explain it. Biological explanation takes many forms. To explain why a frog jumps into a lake, we may invoke a causal chain running from visual information arriving at the retina, travelling to the brain and on to motor nerves which trigger muscular contraction in the frog's legs. We may say that the frog jumps to avoid being eaten by a snake. We may also say that it jumps because of the protein filaments, actin and myosin, in its leg muscles and their capacity to slide across one another. Each explanation is legitimate and self-contained; none can be eliminated in favour of one true causal description. We live in one world, but there are many ways of knowing.
The genetic determinist claim is that the goal of living organisms is the replication of their genetic material-organisms exist as DNA's way of making copies of itself. Further, that DNA forms a "blueprint" which development processes passively follow, albeit in some measure modified by the environment in which this development occurs. DNA explains everything from the egg-laying of the parasitic wasp to the heroism of fire-fighters risking their lives to rescue people from burning buildings.
Such genetic imperialism fails, not only because it ignores the validity of other types of biological explanation, but also because, even in its own terms, it magnifies the role of a single molecule within the self-organising homeodynamic life processes. Organisms have four dimensions: three of space and one of time. DNA is a one-dimensional inert molecule. Far from being self-replicating, it requires an entire cellular orchestra of enzymes, ions and so on to order the synthesis of proteins on which all life depends. During this protein synthesis, DNA sequences are shuffled, carved into sections and reinterpreted to meet cellular needs in such a way that the route from a single piece of DNA-once regarded as "a gene"-is multiply determined. You cannot read the four dimensions of life from the one dimension of DNA. Organisms are not merely created on instruction from their DNA or by selection in their environments, rather they construct themselves. This makes the future, of individuals and species, radically indeterminate. It is our biology, then, that makes us free, though in circumstances not of our own choosing.
Over to you,
Steven Rose
1st November 1997
Dear Steven,
Your characterisation of reductionism is misleading. James Watson's dictum that everything that is not physics is mere social work, and that biologists should try and explain their phenomena in terms of physics, is not a view which the great majority of reductionist biologists would accept. It is not even a view that physicists would agree to. However, we would all accept the fundamental reductionist idea that all phenomena in the universe are controlled by one set of fundamental physical laws. (I cannot believe that you would disagree.) But that is not the same as wanting to explain everything in terms of physics. As you rightly say, there are different levels of organisation-this has long been recognised by biologists. The reductionist approach is to try and account for the phenomena at one level in terms of the properties of the components on the next level down. It is unnecessary to go down to further levels, but there must be nothing in one's explanations that contradicts laws that hold at lower levels-such as the laws of physics. The trick is to find the right level at which to explain the phenomena-no one would attempt to deal with economics in terms of the behaviour of molecules.
The language of cells is, as Sydney Brenner once put it, molecules. Cells are extremely complicated organisms. But our progress in understanding their behaviour has been astonishing. We know that the way to understand cell pro- perties is to understand which proteins are present in the cell and how they interact with each other and other molecules. For instance, we can now understand muscle contraction in terms of the interaction between the proteins myosin and actin-the head of the myosin exerts an active force by moving back and forth when it binds to the actin molecule. We do not really need to understand this in terms of chemistry unless we want to. It is, I have to admit, largely a matter of aesthetic and scientific judgement as to which approach will give you a satisfactory explanation.
I take exception to your claim that reductionism converts dynamic processes into static properties. Please give me a good example. The conduction of the nerve impulse, so elegantly explained in terms of ion movements across the cell membrane, is one of a host of counter-examples. Would you deny that this is both reductionist and dynamic? And what about the way that cells break down food molecules to generate energy-are not those flow diagrams dynamic, and is that not pure and beautiful reductionism?
I cannot understand your hostility to genes. The development of the embryo is determined by the behaviour of its cells-a reductionist position in itself. The behaviour of the cells is in turn largely determined by which proteins are present and how they interact. Which proteins are present is determined by the genetic information in the cells-the DNA. Genes are turned on and off during development, and so the proteins, for which they provide the code, are made or not made. They are not a blueprint, but the programme for development. The difference in development between all animals lies in the genes. Genes are the controlling elements. This is not to say that the machinery they control is not in itself complex. You really must distinguish between the two. The machinery of your computer is complex-but this does not mean that you do not control what you type and print.
I do not know what you are talking about when you introduce seemingly mystical concepts such as "self-organising homeodynamic life processes." We do not need them, because we actually have an extremely good understanding of the principles which govern the embryonic development of organisms at the cell and molecular levels. The early development of the insect embryo-the system we understand best-is a cascade of gene actions. So too is the development of a structure like the limb; the gene products control pattern and form.
All good science is essentially reductionist because it isolates the key components and makes experiments possible.The enemy of reductionism and science is holism.
Regards,
Lewis Wolpert
4th November 1997
Dear Lewis,
Let us agree, first, that no property or process in complex systems contradicts the principles of physics, chemistry and so on; second, that there are many, equally appropriate ways of explaining biological phenomena; third, that reductionist methods and explanations can be perfectly legitimate, within limits; and finally, that we are always interested in finding the right level at which to explain.
But what is the right level? You say it is the one you feel most comfortable with-a somewhat individualistic position, as your preference is presumably determined by philosophy, history and social context (or would you say genes?). This brings you close to those social constructionists whose views on science you deplore. A mere matter of opinion or aesthetics? Surely not. If your right level is different, for personal reasons, from that of a chemist or a physicist, you really have no grounds to dismiss Jim Watson-who is after all a molecular biologist-in such a cavalier manner. I don't think that the right level is to explain alcoholism in Russia or violence in the US in terms of the molecular genetics of individual "alcoholic" or "aggressive" persons-yet significant funding is put into research programmes working along these lines. This, incidentally, is precisely what I mean by converting a process-social violence, for example-into a property, a lump of aggressive biochemistry inside an individual.
As for genes, of course I am not "hostile"; how could any biologist be? What I am hostile to is the quasi-religious view which you espouse when you describe genes as first causes and unmoved movers, a philosophical position which locates you somewhere in the 17th century. Genes in their place, as I explain in Lifelines, means dynamic portions of DNA participating in the orchestrated metabolism of the cell. Molecules-including those of DNA-may be the language of the cell, but understanding the meaning of that language requires knowing the history and culture in which it is embedded. That, for individual organisms, evolving species and-above all-humans, is what homeodynamics is all about.
Affectionately,
Steven
5th November 1997
Dear Steven,
No, I cannot agree with all you suggest. There are not many equally appropriate ways of describing biological phenomena. There is just one way to understand, for example, how proteins are synthesised, and that is in terms of the molecules and their interactions. Tell me, please, another one. (You seem to be remarkably reticent in giving examples to back up your claims.) You say reductionist explanations are alright within limits-such as? And why should I not dismiss Watson: he was not really talking about how one practises science or reductionism, merely praising physics as underlying everything.
But at last you have come clean about what you object to regarding genes. It really relates to your prejudices about human behaviour. You seem unable to tolerate the idea that much of our behaviour is genetically determined. This is not a philosophical or political issue, it is a scientific one. Either intelligence is determined by a large genetic component or it is not. One has to try to find out. Your political prejudices make it hard for you to accept research into this, or into the possible genetic bases of criminality or homosexuality. Your homeodynamics appear to me like a political manifesto. And while your motivation may be noble because you think that such research can be used for what you see as unacceptable social policy, that is politics and has nothing to do with reductionism. Moreover, all knowledge can be put to bad use-just think of fire.
I smell a relativist in your remarks about needing to understand science in its cultural context. Do you seriously believe that science is just a social construct and that there are other ways of describing cell behaviour from a different cultural standpoint? Nonsense. Genes are the controlling elements; this has nothing to do with mystical ideas such as prime movers. Dawkins' concept of selfish genes is excellent and enables us to make sense of numerous complex aspects of biology, not least altruism. You should approve.
Regards-and not really hoping to persuade you,
Lewis
5th November 1997
Dear Lewis,
You are backing away from points which I thought we had agreed on. I used the behavioural example because it flowed from our discussion, but "politics" and "prejudices"-either yours or mine-are absolutely not the issue. What is at stake is the nature and purpose of explanation in biology and how we conceive of living processes. Nor do I believe that science is "just" a social construct, although I certainly do not think that our scientific explanations are free of cultural preferences any more than you do, with your references to which types of explanation make you feel "comfortable." For instance, your assumption about the need to find out whether intelligence is or is not "determined by a large genetic element" presupposes that it is possible to quantify intelligence along a linear scale and measure it by IQ. I do not agree with either of these presuppositions-nor, actually, do many neuroscientists. Of course we all accept that brain processes are necessary for intelligent behaviour and that genes help shape brain processes.
So to your protein example. If we want to know the molecular steps by which the amino acid sequence of proteins is built up from its individual molecular building blocks, we need explanations on the biochemical level, in terms of enzymes and coding sequences (derived, of course, from sequences of DNA). This is a "within level" explanation. If we want to know why these synthetic mechanisms take the form they do, we need to understand first the chemistry of amino acids and proteins, as well as details of their physico-chemical interactions and constraints (the reductionist level), but also the evolutionary steps which have led both to the enzymic mechanisms concerned and to the cellular requirement for this particular protein (among the squillions of other possible combinations). If we want to know why this particular protein and not one of these others is produced at any particular tissue and time in the life of the organism and the cell, we have to ask what part the protein plays in the overall cellular economy. That question cannot be answered from a knowledge of the molecular mechanisms of protein synthesis alone. It is of a different order, but it is, you must agree, as interesting and important a question to biologists because it is one which you and I, in different contexts, have spent part of our own research careers trying to answer.
Yours ever,
Steven
7th November 1997
Dear Steven,
When we talk about cells you seem to be a reductionist just like the rest of us, even going so far as to wish to explain molecular interactions in terms of chemistry. But there is one point where I cannot go along with you. I know it has been said that nothing in biology makes sense unless it is related to evolution. However, your insistence that we need always to understand the evolutionary origin of process seems unnecessary to me. We can talk about-and understand-cellular mechanisms without understanding their evolutionary origins. This is particularly true in medicine when we want to understand the origin of a disease. We can try to unravel the molecular mechanisms of, say, Aids, without an evolutionary perspective. Our area of disagreement remains, in fact, in the problem of human behaviour. To talk about the process of intelligent thinking is not to reify it. I do not think that you would deny that some people are better at maths than others, and that some are very gifted. It seems logical to ask to what extent such differences are caused by genetic or other factors. Intelligence is difficult to measure, but that should not stop us trying to do so-provided that we recognise the limitations of our research. The evidence for a genetic contribution to what we think of as general intelligence is now overwhelming. But I must admit that an illness such as depression does raise for me some difficult reductionist problems. I still do not know what I would regard as a satisfactory explanation, or at what level of organisation I would look for it. It is not easy to sort out the relation between the biological processes in the brain and the psychological ones. It may be necessary to link, for example, the emotions associated with depression with the action of serotonin on complex networks of neurons in the brain. At the moment we cannot even easily explain the behaviour of quite simple neural networks in reductionist terms. However, in the end a reductionist explanation will remain the most satisfactory.
Regards,
Lewis
