This article was initially published, in French, in the book Espèces et éthique. Darwin: une (r)évolution à venir (“Species and Ethics: Darwin, an upcoming (r)evolution”), ed. Tahin-Party, 2001.
Read time: 48 min.
Abstract: In defence of his selfish gene thesis, Richard Dawkins has given two parallel versions: that of a redefinition of the term “selfishness” and that of its use as a metaphor. Neither stands up to scrutiny. In particular, they can neither help to understand the existence of non-coding DNA, nor validate the view of organisms as “survival machines” for genes.
The selfish gene thesis is therefore flawed by its attribution of finality, or pseudo-finality, to genes. It turns out, however, that hidden in this failure is an important truth that he himself does not make clear enough: if genes are not selfish in the sense of “caring only about themselves”, they are indeed selfish in the sense of "not caring about others". Dawkins' thesis is in fact opposed, not to hypothetical biologists who, following Darwin, have well understood the absence of finalism in natural evolution, but instead to that great mass of biologists who persist, in spite of Darwin, in interpreting natural processes as driven by a higher order finality and want that genes exist for the good of the organism, the species or even the entire ecosystem. The scandal of Dawkins was not in his finalism, but in the poverty of his almost infinitely fragmented finalism. In this sense, Dawkins did a very useful work, even if he stopped along the way.
Among the contemporary currents that claim their attachment to Darwinism, there is sociobiology. The approach is controversial, in some countries more than in others; in France, in particular, it carries a whiff of sulphur – that is, it is allegedly far-right.
However, it may serve to look beyond odours. There is too much of a reflex of systematic suspicion to look for hidden intentions behind any thesis, to the detriment of examining the thesis itself. Regardless of the question of possible ideological interests that sociobiology may serve or not, it is useful to ask: is what sociobiology says true or false? And first of all: what does sociobiology say?
I will not claim to have answered such a vast question, but only to examine the meaning that can be given – or not – to a key expression of the theses of Richard Dawkins, one of the two most prominent figures of sociobiology1, which makes up the title of the book that rendered him famous, The Selfish Gene2, and to reflect more generally on the role of finalism in Dawkins, and in his colleagues – whether sociobiologists or not.
To avoid misunderstandings, let us first clarify a point of vocabulary concerning the identity of the “gene” object as discussed in Dawkins. This term may suffer from the same ambiguity as, for example, the word “book”: it is not of the same book that we say that it is on the table and that it is successful. The book on the table is a concrete object, also called a “copy” of the successful book; the latter, on the contrary, is an abstract object, defined (for simplicity) by its text, by a certain sequence of characters.
Dawkins often speaks of genes that “survive” or “perpetuate”; they do not do so as concrete objects, however; the atoms they are made of are dispersed from one generation to the next. It is as identical copies that the gene perpetuates itself, that is, as a “text”, a structure. The text of a gene is its nucleic acid sequence; it “survives” as long as this sequence continues to exist in identical form from one generation to the next.
1. Genes that are selfish by redefinition?
The title of Dawkins' book suggests a simple formulation for his thesis:
Genes are selfish.
To this, the philosopher Mary Midgley replied:
Genes cannot be selfish or unselfish, any more than atoms can be jealous, elephants abstract or biscuits teleological. This should not need mentioning, but Richard Dawkins's book The Selfish Gene has succeeded in confusing a number of people about it (...)3.
This rebuttal appears to be irrevocable: Dawkins makes a category error, applying to genes a concept – selfishness – that is not intended for them. Dawkins, however, protests that he has been misunderstood, and invokes the redefinition of selfishness made in the first pages of The Selfish Gene:
We define altruism and selfishness in purely behaviouristic ways: “An entity ... is said to be altruistic if it behaves in such a way as to increase another such entity's welfare at the expense of its own. Selfish behaviour has exactly the opposite effect. “Welfare” is defined as “chances of survival”, even if the effect on actual life and death prospects is ... small ... It is important to realize that the above definitions of altruism and selfishness are behavioural, not subjective4.
Such a redefinition seems apt to justify a strictly literal use of terms like “selfishness” and “altruism” in relation to genes and other entities that are not capable of being selfish or altruistic in the usual sense. In short, the redefinition negates any previous meaning of the word, and Dawkins might as well have said “glik” and “glok” instead of “selfish” and “altruistic”, and entitled his book “The glik gene”. But he did not.
This is because this redefinition does not actually cancel out the previous meaning of the words concerned, but rather operates on their meaning in a collective transposition. In the passage quoted, Dawkins explicitly redefines three terms: altruism, selfishness and welfare. In each case, the operation consists in subtracting their subjective component: altruism and selfishness are usually defined by the will and the behaviour resulting from this will; in Dawkins' definition, it will be by the behaviour alone5 . Survival, an objective concept, is normally only a condition of welfare, a subjective concept; here it will be its whole6.
This subtraction preserves the articulation that links the various terms. Altruism and selfishness are antonyms, both before and after the redefinition, and remain in the same relationship with welfare, itself redefined. Moreover, in Dawkins' texts, this articulation implicitly extends to a collection of other terms, to the whole network of those with which the notions of altruism and selfishness are usually associated: struggle, competition, cooperation, manipulation, goals, success, attempt to, at the expense of, aspire to, and so on – and not least to the little word “for”, frequent support of discreet finalism. The whole of this terminological network refers in everyday language to subjective phenomena, and each of its members would therefore have to be explicitly redefined before Dawkins could legitimately claim to use it non-subjectively in conjunction with his non-subjective version of altruism and selfishness.
Yet he does not take the trouble to do so, and despite this, his text remains readable. This is because we have understood, without being told, the procedure to be applied whenever necessary: to keep only the objective residue of the meaning of words. So when we read Dawkins we constantly apply this recipe – or think we are applying it. For the operation can be tricky if we try to do it for real.
Let us consider the case of “struggling”, an activity Dawkins attributes to genes, or, as in the following passage, to their “primordial soup” ancestors, the early replicators:
The next important link in the argument (...) is competition. (...) There was a struggle for existence among replicator varieties. They did not know they were struggling, or worry about it; the struggle was conducted without any hard feelings, indeed without feelings of any kind. But they were struggling, in the sense that7...
In what sense could they struggle? What is the objective residue of the verb “to struggle”? It is not difficult to identify it in other concrete cases; we know, for example, the movements of two struggling boxers, and could call these movements themselves “struggling”. However, this redefinition would be specific to boxers. How can it be done in the case of genes? If some genes were conscious and others not, the former could occasionally struggle in the full sense of the word and their movements could serve as a model for a non-subjective definition of struggling in the latter. The problem is precisely that we lack such a model!
Here is Dawkins' answer:
But they were struggling, in the sense that any miscopying which resulted in a new higher level of stability, or a new way of reducing the stability of rivals, was automatically preserved and multiplied.
Yet to say that the replicators were struggling should at least imply that they were doing something. In this mere description of the situation – “any miscopying... was automatically preserved and multiplied” – we do not see what the behaviour is, what are the subject, verb and complement. Can the replicators be the subject of an action that would be the miscopying? Or are they acting by just being stable?
Dawkins says neither. He begins, however, by indicating that he has applied the general recipe, that of determining the objective residue: the first use of the term “struggle” precedes its redefinition, the passage from one to the other being made by subtracting the subjective elements (“They did not know they were struggling...”). Thus he inserts an additional word – “to struggle” – into the redefined terminological network of altruism and selfishness, by a plain assertion, without describing its content.
Why does he not indicate its content? In particular, why does he not tell us that the replicators were fighting by miscopying? Is this not how they can beat their “rivals”? The problem is that such a copying error, for the replicator, cannot be viewed as a behaviour that increases its chances of replicating: on the contrary, it reduces these chances to zero!
To see this, it is easier to examine the selfish gene thesis from another angle, that of a metaphor. For the moment, let us note that Dawkins cannot use his redefinition of selfishness and altruism to legitimise the literal use he claims to make of these terms and the entire terminological network to which they belong.
2. Or selfish by metaphor?
Despite his response to Midgley, it is not the literal use of a redefined concept of selfishness that Dawkins most often puts forward. Rather, selfishness is attributed to the gene by metaphor:
Throughout this book, I have emphasized that we must not think of genes as conscious, purposeful agents. Blind natural selection, however, makes them behave rather as if they were purposeful, and it is convenient, as a shorthand, to refer to genes in the language of purpose. (...) the idea of purpose is only a metaphor, but we have already seen what a fruitful metaphor it is in the case of genes. We have even used words like “selfish” and “ruthless” of genes, knowing full well it is only a figure of speech8.
So do we have a redefinition, or a metaphorical use of these terms maintained in their common meaning? The two procedures amount in fact to the same, to a large extent, which explains why Dawkins can switch effortlessly from one to the other. Redefinition has the advantage of allowing for literal use and thus making the statements appear rigorous; a metaphor, instead, appears as a mere “figure of speech”, that is, as a suggestive expression that is not to be taken seriously, “literally”.
Metaphor consists in building what I will call the world of “as if”: genes are not conscious agents, but behave, says Dawkins, as if they were. In the imaginary “as if” world, genes are conscious agents; they behave in a certain way. The real world is identical to the “as if” world, except that in it the genes are not conscious; their behaviour, however, is the same. In short, one moves from the world of “as if” to the real world by subtracting subjectivity and keeping the objective residue – in other words, by carrying out the same operation as in the redefinition of the word “selfishness” in a purely “behavioural”, non-subjective way, and the concomitant redefinition of the whole terminological network linked to it.
The metaphor seems elusive because it gives the feeling of literally asserting nothing. What point would there be in refuting – or confirming – statements that their own author does not take seriously? We see, however, that the use of the metaphor of the selfish gene is itself a non-metaphorical statement: it asserts, literally, that things happen as if genes were selfish conscious agents. We can therefore test this metaphor as a literal statement. We must ask: how would things be in a world where genes were selfish conscious agents, with only one idea in mind, that of self-preservation and multiplication?
So let's put aside our rationalist inhibitions and dive into the metaphor: let's imagine that genes are conscious. Better still, let's slip into the shoes of one of them.
I am a gene, and my goal is to preserve myself and multiply. I want the exact sequence of nucleotides that defines me, that is me, to be present in as many copies as possible in tomorrow's world.
What can I do to further my goal? What are my means? On what does my success or failure depend? It is the properties of the body that carries me that will determine the number of copies of myself in the world of tomorrow. There are also other factors, such as the environment, and of course plain chance, but I have no direct influence on these; it is only through the physical and behavioural characteristics of my body that I can hope to act towards my ends.
So I will make this body, for example, strong; or perhaps it will be to my advantage to make it weak but lighter. Or perhaps I will make it aggressive, or instead gentle and kind, according to what serves my interests. In any case, I will shape this body to maximise, in the given circumstances, the number of genes identical to myself in the future world.
Let's say, for example, that I am a gene that determines whether my body will be strong or weak. Suppose it is advantageous for my preservation and propagation that my body is strong. So I will make it strong...
At least, I would like to. It is not quite true that I determine whether the body will be strong or weak. I don't determine whether, but only that, the body will be strong, or that it will be weak. If I am a certain sequence, the body will be strong. Good for me. If I am another sequence, the body will be weak. So much the worse for me. I have no choice. I am what I am, that is, a sequence and not another, of nucleotides. If I “determine” any property of my body, it is so to speak passively. My desire to preserve and propagate myself may influence my dreams, but not my actions. If I am a gene for weakness, I can do nothing about it.
I can do nothing about it... except by mutating. That's my means of action! I, a gene of weakness, will mutate and become, in the next generation, a gene of strength. Thus – O joy! – I will have achieved my goal, and countless copies of myself will pour into the world.
O joy? No, O cruel fate! For what will pour forth will not be me. I am a sequence of nucleotides, not another. If I mutate, I will no longer be me.
In short, my will to preserve and propagate myself can have no effect; it is my identity, not my will, that entirely determines the influence I have on the body. To change this influence, I would have to mutate, and thus disappear as myself – which is precisely what I do not want.
In a trivial sense, Dawkins is right: the world is indeed as if genes had the will to preserve and propagate themselves. This is true, however, only because such a hypothetical will would have no effect. The metaphor is valid insofar as it is devoid of substance. The selfishness attributed to the gene cannot be the cause of selfish behaviour, in the redefined sense of the term, that is, of behaviour that increases the gene's chances of survival at the expense of others. For we see that this behaviour is fixed, inherent in the identity of the gene (its nucleotide sequence), and can only vary at the cost of it losing its identity.
3. Selfishness by redefinition, again
Perhaps we could at least assert that in fact this behaviour is such that it increases the gene's chances of survival at the expense of the chances of other genes? This brings us back to the thesis of redefined selfishness. I criticised above Dawkins' use of the word and the global transposition he makes of the terminological network to which it belongs; but can we not at least retain the basic assertion: that genes are selfish, in this redefined sense, that is that their behaviour does, in fact, increase their own probability of survival and decrease that of other entities of the same type?
The “other entities of the same type” will typically be the other alleles, that is, the other genes that could occupy a given location, or “locus”, on a chromosome. In the imaginary example mentioned above, we have two alleles, one whose property is to make the body strong and the other to make it weak. This property is immutable since it results from the very identity of the genes involved, but let us assume that it can be given the status of a behaviour. Can we say that this behaviour actually increases the chances of survival of the allele at the expense of the other allele?
“To increase” involves a comparison, in this case with one or more other behaviours. But what will these other behaviours be, in the case of our two alleles?
The first answer that comes to mind is that the behaviour of each of our alleles must be compared to that of the other. To say that they are selfish is therefore to say of each that its behaviour gives it a greater chance of survival than the other gets from its own behaviour. But we cannot say this simultaneously for both alleles. One at most can be selfish in this sense, the other being altruistic. In our example, the gene that makes the body weak is altruistic, since it decreases by its behaviour its chances of survival and increases those of the other gene; only the latter, which makes the body strong, is selfish.
“Some genes are selfish”, Dawkins should thus tell us. But how many are? Since selfish genes, by definition, survive longer than altruistic ones, won't the latter quickly disappear? Let's assume that this is the case; that in our example only the gene that we called selfish, the one that makes the body strong, remains in the running. If the same thing happens at every locus, can we not say that all genes are selfish?
Can we not then say that all genes have a behaviour that increases their own chances of survival and decreases those of other genes? But then, what other behaviours should we compare these to? If we assume that altruistic genes have disappeared, we lose the other term of our comparison.
All that we would have left would be to postulate that the comparison must be made not to other existing alleles, but to all conceivable alleles for the locus in question. The gene is selfish if and only if its behaviour gives it a greater chance of survival than any other possible behaviour. To say that genes in general are selfish would then be to say that every existing gene is the one that has the greatest chance of survival of all conceivable genes at that locus.
Indeed, many have believed this to be the meaning of Dawkins' theses. Dawkins admits, at the beginning of Chapter 3 of The Extended Phenotype9, that he finds some merit in the so-called adaptationist attitude, “that approach to evolutionary studies which assumes without further proof that all aspects of the morphology, physiology and behaviour of organisms are adaptative optimal solutions to problems”10. The entire Chapter 3, however, is devoted to listing the reasons why actual “adaptations” are precisely not optimal. The genes are not, in the general case, those that have the greatest chance of survival relative to all conceivable genes for the same locus.
In Dawkins' redefined sense, many genes are selfish, but of a limited selfishness; that is, they have a behaviour (properties) that causes them to survive more than would many (but not all) conceivable alternative behaviours. They are often also selfish in relation to existing genes, when the less “efficient” allele(s), which are therefore altruistic, have not yet disappeared – if they are indeed destined to disappear. For selfishness and altruism are in this sense as much properties of the environment as they are of the genes, and whenever the environment changes, a given gene from altruistic can change to selfish11.
In any case, the selfishness attributed to genes does not justify the wholesale recovery that Dawkins tends to make of the entire terminological network with which the word is associated in its everyday sense, including terms such as “competition”, “struggle”, etc. Nor does it provide a basis for a properly functioning metaphor.
The concept of the selfish gene seems to boil down to the observation that genes frequently exist that, in a given environment, preserve and perpetuate themselves more than others would do in their place. In short, it seems useless, when it is not misleading.
4. Carefree DNA
Untranslated DNA represents an extreme example of the difficulty inherent in the metaphorical selfishness thesis12 . Yet Dawkins seems to believe that it provides, on the contrary, its resounding confirmation.
Here is the problem as he poses it:
(...) it appears that the amount of DNA in organisms is more than is strictly necessary for building them: a large fraction of DNA is never translated into protein. From the point of view of the individual organism this seems paradoxical. If the “purpose” of DNA is to supervise the building of bodies, it is surprising to find a large quantity of DNA that does no such thing. Biologists are racking their brains trying to think what useful task this apparently surplus DNA is doing. But from the point of view of the selfish genes themselves, there is no paradox. The true “purpose” of DNA is to survive, no more and no less. The simplest way to explain the surplus DNA is to suppose that it is a parasite, or at best a harmless but useless passenger, hitching a ride in the survival machines created by the other DNA13.
We have seen that the selfish will metaphorically attributed to the gene is powerless to act on the world. The gene does, however, normally have an “action”, a phenotypic effect that can at least pass for the sign of such a will. Here this action is entirely lacking: yet the gene survives, and Dawkins persists in relating its survival to its will. It is understandable that some people, faced with such “Darwinian” explanations, might find less extravagant the account given in Genesis – “And God said, Let the earth bring forth the living creature after his kind (...): and it was so.”14. God at least takes the trouble to say something in order to act on the world; the gene needs no more than to will.
What is to be concluded from such a patent argumentative gap? I believe that at least one negative fact emerges here: the role played by the gene's selfishness in Dawkins' perspective is not the one that he himself suggests; its selfishness has no value as an explanation, either real or metaphorical, of either its behaviour or its survival. In particular, a closer reading of the end of the above passage reveals that it is not quite to the will of the DNA – to its “true ‘purpose’” to survive – that is given the explanatory role. This role is rather assigned to a certain ontology of the DNA: to its nature as a “parasite”, a “useless passenger”. A negative, hollow ontology – as we will see.
5. The “survival machine” and the hard-working gene
Not all strands of DNA allow themselves to float carelessly along the winds of natural history. In contrast to this negative ontology of carefree DNA, we find in Dawkins a recurring discourse on the nature of living beings themselves, a discourse in which the gene appears as an agent in a great collective work.
In the title of the first chapter of The Selfish Gene, Dawkins asks the question: “Why do we exist?” Here is his answer:
Now [the replicators] swarm in huge colonies, safe inside gigantic lumbering robots, sealed off from the outside world, communicating with it by tortuous indirect routes, manipulating it by remote control. They are in you and in me; they created us, body and mind; and their preservation is the ultimate rationale for our existence. They have come a long way, those replicators. Now they go by the name of genes, and we are their survival machines15.
In another text he makes this statement the very heart of Darwinism:
By far the most important thing that the Darwinian hypothesis tells us about ourselves concerns that fundamental question which until now has been answered mostly in religious terms: why do we exist? The answer to this fundamental question is that our body is a mechanism for the preservation and propagation of our genes. This conclusion cannot be doubted by anyone who has seriously thought about the problem16.
“Why do we exist?” The word “why” can ask about the cause (“Why did the hoover stop working?”), or about the aim (“Why?” = “for what purpose?”). It is in the latter sense that Dawkins' question should be understood since he does not answer “because of...” but “for...”: for the preservation and propagation of our genes. “For” implies an end; hence subjectivity17.
It is in this context that the lines I quoted above concerning the “competition” and “struggle” between replicators – a struggle that is “conducted without (...) feelings of any kind” – are situated. We see that Dawkins cannot be interpreted consistently here. If we take seriously his repeated assertion that he attributes to genes only the objective residue of words, whether through their redefinition or through their metaphorical use, we cannot take seriously – literally – that other repeated assertion which is his answer to the question of “why we exist”. We cannot be literal machines created by genes for an end if those genes are not otherwise animated by any feeling, by any will to achieve such an end. This inconsistency in Dawkins is all the more serious because, as I see it, the fundamental meaning of Darwinism lies precisely in the assertion that it is possible to build an entirely causal explanation of evolution.
This idea of the “survival machine” brings, I think, another inconsistency into Dawkins' thinking. We have seen that the gene that is at stake is not the material piece of DNA, but a sequence of nucleotides. The gene “survives” as long as this sequence lasts, copied identically from generation to generation.
Thus defined, Dawkins tells us, genes are forever:
Individuals are not stable things, they are fleeting. Chromosomes too are shuffled into oblivion, like hands of cards soon after they are dealt. But the cards themselves survive the shuffling. The cards are the genes. The genes are not destroyed by crossing-over, they merely change partners and march on. Of course they march on. That is their business. They are the replicators and we are their survival machines. When we have served our purpose we are cast aside. But genes are denizens of geological time: genes are forever18.
Forever? Not really: because of mutations, but also because of the crossing-over that, in sexual organisms, occasionally cuts the gene itself in two pieces and exchanges one of these with the corresponding piece of another allele. Dawkins' discussion of this last point shows that the gene is in fact a rather short-lasting entity19 . This does not prevent the same Dawkins from describing us as survival machines patiently developed over several billion years of evolution by the genes that were already floating in the “primordial soup”. Machines for the survival of genes... which, however, as precise sequences of nucleic acids, have in their great majority not survived. And even if they were all still there, the “survival machines” could not be of their making. The gradual development of these machines depends on the continual appearance of new genes, which could not – without being very altruistic – build these machines as devices for the survival of the older ones.
But what did this image of the survival machine developed by the primitive “replicators” suggest to us? “They have come a long way, those replicators”, Dawkins told us. To come a long way is to evolve, yet without losing one's identity. The image of the “survival machine” can only be understood if we identify the gene with an entity capable of innovating, and therefore of evolving without disappearing; thus neither with a concrete molecule, nor with a precise sequence of nucleotides, but with a lineage.
We have seen that it is impossible to make the metaphor of the selfish gene work because of the lack of means that it suffers. The only possibility it has to “change its behaviour” is to mutate, but in mutating it disappears. But this is only true as long as we define the gene as a specific sequence of nucleotides; if instead we define it as a lineage, it survives mutation. Here will then lie its means of action. The gene for weakness will mutate to become a gene for strength, without disappearing; on the contrary, it survives, longer and better than it would have done without the mutation!
Such a “solution”, however, poses various problems, not least of which is the broadening of its identity, which risks rendering meaningless the selfishness we attribute to it. For the lineage is not just a line, but a sequence of branches. If the gene retains its identity when it is mutated, it will be the same object not only as its father-gene, but also as all its brother-genes, nephews, uncles and so on. Being selfish, it will care only about itself, but this “itself” will be very extensive and may even include all the alleles with which Dawkins sees it competing!
In his lyrical statement such as the one above about the nature of our being, Dawkins tends to abandon his usual position, centred on the smallest units, and adopt instead a “holistic” language. “That is their role” he says of genes that “march on”. In what play have they received such a “role”*? Of what city are they “citizens”? Genes remain protagonists but as a collective entity, architects of life, of survival machines. I think it is in these passages that Dawkins is most seductive to an audience hungry for pre-Darwinian finalism and most removed from the aspect of his work in which he is original and authentically Darwinian. It is to this aspect that I will now turn.
The Plainly Selfish Gene
Midgley, as we have seen, criticises Dawkins for making a category error in speaking of selfish genes – “Genes cannot be selfish or unselfish, any more than atoms can be jealous...”. This objection seems final if one uses the word “selfish” literally – not metaphorically – and in its usual – not redefined – sense, to genes as we understand them, aggregates of atoms devoid of desires and thoughts.
But are we sure? Let us examine the common meaning of selfishness, through the following two usually equivalent definitions:
— To be selfish is to not care about others (s1)
— To be selfish is to care only about oneself (s2)
Formally, (s2) adds a positive condition to the purely negative content of (s1): to be selfish in the sense of (s2), an entity must not care about others (as in (s1)), but must furthermore care about itself.
Now, (s1) is entirely true of genes: they do not care about anything, and therefore, in particular, do not care about others. Genes are selfish in the sense of (s1), literally, with no need for metaphor or redefinition.
They are not, however, selfish in the sense of (s2): for they care neither about others nor about themselves. The definitions (s1) and (s2) can be equated in everyday life because the entities to which they are typically applied – adult, conscious human beings – always care at least about themselves; indeed, they care all the more about themselves the less they care about others. The additional condition required by (s2) is always fulfilled and is fulfilled all the more so as condition (s1) is fulfilled.
This is not true in the case of genes, which, unlike us, are accomplished masters of the perfect mental void. They are supremely selfish in the (s1) sense, without being selfish in the (s2) sense.
Let's go a little beyond this formal game. In (s1) the expression “care about others” actually packages a whole series of finalistic concepts: aiming at the good of others, acting for others, being useful to others, existing for the service of others, etc. Thus, the entity that does not aim at the good of others, that does not act for their good, that is not useful to them, that does not exist to serve them, is selfish in the (s1) sense. It may seem trivial to assert that the gene does not care about others, since it does not care about anything; the proposition becomes less trivial if we formulate it in the following way:
Biologists are wrong to postulate that a stable gene must exist for the good of the whole organism, or of the species or of any group to which that organism belongs, or of the set of other genes, and, more generally, that it must perform a function or possess a utility for a whole greater than itself.
Let us return to the passage already quoted concerning untranslated DNA:
(...) a large fraction of DNA is never translated into protein. From the point of view of the individual organism this seems paradoxical. If the “purpose” of DNA is to supervise the building of bodies, it is surprising to find a large quantity of DNA that does no such thing. Biologists are racking their brains trying to think what useful task this apparently surplus DNA is doing.
These biologists, Dawkins tells us, are wrong. DNA is selfish, in the (s1) sense: its existence is not in the service of anything but itself.
But does DNA serve itself? Does it exist for itself? This is where Dawkins seems to slip from (s1) to (s2):
But from the point of view of the selfish genes themselves, there is no paradox. The true “purpose” of DNA is to survive, no more and no less.
This attribution of selfishness in the sense (s2) comes, however, after the formulation of a problem that is not that of Dawkins, but of biologists who would like DNA to be at the service of the whole organism. It leaves the field open for the assertion:
The simplest way to explain the surplus DNA is to suppose that it is a parasite, or at best a harmless but useless passenger, hitching a ride in the survival machines created by the other DNA.
We appear here to be back to the classical pattern of essentialist thought that explains things not by a sequence of causes and effects, but by the ontology of objects. From a causal point of view, this is a non-explanation, but it is enough, since the problem itself – the “paradox” – has been dismissed by the previous sentence. As for the ontology that is asserted, it is minimalistic: the nature of DNA is simply to exist. This nature is fully realised, certainly, in all existing genes, and it only serves to explain this existence because it replaces another ontology that made this existence problematic.
In this argument, the objective attributed to DNA has no other function than to serve as a sort of verbal counterpoint to the classical purpose of “supervis[ing] the building of bodies”; as if there was a box, labelled “the objective of DNA”, that had to be filled, and that if one objective was dismissed it was necessary to fill the box with another, even if this other was devoid of substance. Formally, it is thus selfishness (s2) and not (s1) that is attributed to the gene. But the gene's selfish dream is not to eat strawberries or go out clubbing! It just wants to exist. Again, we have an empty desire, that is fulfilled in all the genes that exist, and a selfishness (s2) that reduces, in its substance, entirely to (s1).
So what Dawkins is defending is a counter-ontology, rather than an ontology properly speaking, and a counter-finality of the gene rather than a finality. The metaphor itself, the one that has shown itself incapable of functioning as such, can in turn be seen as a counter-metaphor. The real metaphor is that of the altruistic gene, the gene at the service of the organism or the species. For if genes were altruistic, what would happen? We have seen that the gene for weakness was powerless to act, its only means of action, mutation, being contrary to its purpose, its own survival; but if we give it an external purpose, the contradiction disappears. It can mutate and become a strength gene, for the good of the organism; its disappearance as a precise sequence of nucleotides does not matter to it. It continues to exist as a faithful servant of the organism. It appears that it is in fact this metaphor that is being asserted when it is said that evolution, which is that of genes, is in the service of the organisms or the species.
That the metaphor of the altruistic gene makes sense does not imply that it is correct as a description of reality. What Dawkins argues is that it is not correct; that is, that things, in reality, are not as if the genes were altruistic. Altruistic genes could act, but the result of their action would not be what we observe in reality.
To this metaphor of the altruistic gene, authentic as a metaphor but false in its discourse on reality, Dawkins opposes a false metaphor, built as a sort of inverted image of the first. But the inverted image of a metaphor is not itself a metaphor; we have seen that it does not work, or works only trivially. What matters, though, is not for it to work, but for it to take the place of the metaphor that it aims to counter.
7. Grammatical traps
Why does Dawkins feel the need to create these empty shells? Why build a counter-ontology, a counter-finality and a counter-metaphor, rather than simply throw out the ontology, finality and metaphor? It may well be at least in part for a simple question of rhetorics. It is easier to replace certain elements in a discourse with others – be they empty – than to change its structure.
Thus to an assertion like:
The purpose of DNA is to supervise the construction of bodies.
it is easier to oppose:
The purpose of DNA is not to supervise the construction of bodies. (1)
DNA has no purpose.
But in (1), “the purpose of DNA”, posited as the subject of the sentence, acquires reality by that fact alone; and as soon as we say what it is not, we must also say what it is.
Perhaps Dawkins could have escaped this kind of grammatical trap if the same constraint were not found at the heart of traditional Darwinian vocabulary. I have spoken at length in another article about the finalism inherent in the phrase “natural selection”20 . The same is true of “adaptation”. Dawkins talks about it in a very ingenuous way:
In 1957, Benzer argued that “the gene” could no longer continue as a single, unitary concept. He split it into three (...). I have suggested adding a fourth unit, the optimon, the unit of natural selection (...). Independently, E. Mayr (...) coined the term “selecton” to serve the same purpose. The optimon (or selecton) is the “something” to which we refer when we speak of an adaptation as being “for the good of” something. The question is, what is that something; what is the optimon?
The question of what is the “unit of selection” has been debated from time to time in the literature both of biology (...) and philosophy (...). I agree with Williams (...), Curio (...) and others that there is a need to develop a serious science of adaptation – teleonomy as Pittendrigh (...) called it. The central theoretical problem of teleonomy will be that of the nature of the entity for whose benefit adaptations may be said to exist21.
Here again, Dawkins could have noted that evolution does not have to be for the benefit of anything; but he chooses instead to ask for the benefit of what it is, because “natural selection” implies a unit of selection; and, similarly, the word “adaptation” governs “for the benefit of x” and thus implies the existence of such an x, an “optimon”. Adaptation takes place through genetic modifications. If these are not for the good of the universe, nature, God, humanity, the species, the group, the organism, the cell, or the genome as a whole, we are left with assigning them to the good of the gene itself.
In short, (s1) changes into (s2) following a logic that Dawkins did not invent. His positive attribution of a will to the gene usually operates at a purely verbal level and does not influence the substance of the reasoning, at least when he remains faithful to his good intentions: “(...) we shall always keep a sceptical eye on our metaphors, to make sure they can be translated back into gene language if necessary”22 , that is, into the language of simple causality.
8. One abstract elephant may hide another
Behind the apparent debate – whether genes care or care not about themselves (literally, metaphorically, by redefinition, etc.), – there is a parallel but reversed debate: whether they do or do not care about others. Midgley begins her paper by arguing against the selfish finalism attributed to genes, but most of the rest of her text is devoted to insisting, not on the absence of goals in evolution, but instead on their multiplicity:
Organisms are selected as individuals, but what are they selected for? The term “select” leads people to hope for a simple, positive answer to this question, a single, isolable purpose. (...) It is probably necessary, for evolution as for economics, to think not of one single aim, but of a number which converge (...)23.
Another anti-sociobiologist – Stephen Jay Gould – attributes this finalism to individual organisms:
Natural selection dictates that organisms act in their own self-interest (...) They struggle relentlessly to increase the representation of their genes at the expense of their fellows. We have discovered no higher principle in nature than this stark reality24.
We have seen no great rush to retort to Gould that an oak tree can no more struggle to increase the representation of its genes than an atom can be jealous, an elephant abstract or a biscuit teleological!
In a recent book25, Jean-Jacques Kupiec and Pierre Sonigo also criticise the finalism attributed by Dawkins to genes, but this time to locate this finalism at the level of the individual cell:
There is therefore an ecosystem in each of us, composed of billions of tiny microscopic animals, which we call our cells. They live for themselves, not for us.
Further on, on the subject of white blood cells:
Can we similarly imagine that the white blood cell is only maximising its own profit and does not care about our health? that our health is only the effect of an invisible hand generated by the interactions of selfish cells?
And viruses too want to perpetuate themselves, at least “figuratively”:
Viruses have only one thing in mind: to perpetuate themselves. Of course, the expression is meant figuratively. Viruses have no heart or head, let alone a mind26.
In sum, while Dawkins convincingly shows that the gene does not care about others – and states that it therefore cares about itself – his opponents convincingly show that the gene does not care about itself – and conclude that it “cares” about others, that is, obeys a higher order finalism.
It would be wrong, however, to simply dismiss Dawkins and his colleagues as equally guilty of naturalistic animism. To attribute a finalistic behaviour to the whole organism, for example, is to imbue all its parts, each of its organs, cells and genes, deemed to be each in the service of the whole, with a derivative finalism. Finalism spreads from top down. But it doesn't spread from bottom up.
For to assert that each part acts for its own good does not generally imply that the whole acts for some “good of the whole”. What makes finalism interesting as a mode of explanation is that it allows us to dispense with the often complex and always somewhat “accidental” sequence of mechanical interactions that is the substance of any causal explanation. Instead, the behaviour of the entity that has a goal is directly explained by that goal. But when several entities, each with its own purpose, are brought together, the way in which their individual behaviours combine takes on again the appearance of a sequence of mechanical interactions which, in general, cannot be simplified by any collective purpose. This fact often seems paradoxical to us – we wonder about wars that nobody wanted; but it is very real, as the classic “prisoner's dilemma”27 shows.
It is at the lowest level, that of the gene, that Dawkins places finalism. One of his key points is that this is not the same as attributing it to a higher level. For example:
(...) if adaptations were designed by God, He might have designed them to benefit the individual animal (its survival or – not the same thing – its inclusive fitness28), the species, some other species such as mankind (the usual view of religious fundamentalists), the “balance of nature”, or some other inscrutable purpose known only to Him. These are frequently incompatible alternatives. It really matters for whose benefit adaptations are designed. Facts such as the sex ratio in harem-forming mammals are inexplicable on certain hypotheses and easily explicable on others29.
Thus, in elephant seals, as many males as females are born, which appears to be contrary to the “interests of the species”: for only a small minority of males reproduce, each with his “harem” of females. Meanwhile, the others, who will never have offspring, consume and “waste” the common food resources. The “interests of the species” would be better served if most males were simply not born. Yet they are born, and this sex ratio, which is close to unity, can be explained by focusing on the reproduction rate of a mutant gene that would cause it to deviate from this value30. Dawkins speaks of this result in terms of an “optimisation strategy” at the gene level, that is, in finalist terms, but insists on the absence of optimisation, and therefore of finalist behaviour, at higher levels.
It seems to me that, paradoxically, a large part of the criticism levelled at Dawkins' theses is motivated not so much by his attribution of finalism to genes as by his refusal to postulate the “organic” and “holistic” functioning of a higher level into which this multitude of fragmentary selfishnesses would harmoniously integrate. We are still living in this dream, curiously present in practically the same form in Christianity, in the liberalism of the “invisible hand” and in Marxism, of an automatic agreement between individual and collective interests, an agreement that would ultimately allow us to dispense with ethics, retaining as sole ethical prescription that of “well-understood selfishness”31. If Dawkins too adhered to this scheme, it might indeed lead him to want to deduce from the selfishness he attributes to genes a warlike and cynical ethic at the individual and/or social level, in non-human animals (which is what almost everyone does), and in human beings (which would be a great fascistic crime). But Dawkins does not adhere to this scheme, whatever his opponents32, who are the victims of their own projections, and some of his “supporters33” may say.
This does not mean that Dawkins is blameless on this point; far from it. For the tendency, still universally prevalent today among both biologists and the general public, to attribute a finality to biological entities, especially higher ones, does not only correspond to the ideological will to ensure the perpetuation of a speciesist, sexist, racist and classist “natural order”; it also constitutes an attempt to explain the phenomena of life.
Life is the unique seat of true ends: among the living organisms are animals, and animals alone are sentient, are the seat of subjectivity and therefore of purpose. But in addition to this, there are many phenomena in life that seem to be organised for an end, without our being able to identify the organising will. The eye seems not only to see but to have been made to see. A further step is to view the totality of the living world as the result of one single will; whether it is that of a supernatural entity or corresponds to an intrinsic finalism of nature.
That it is tempting to explain life as the result of a will is undeniable, at least as a historical fact, since human beings have long sought such explanations and thought they had found them. Even today, the vast majority of humans, and not only those who follow one of the three monotheistic religions, adhere more or less consciously to this scheme, as witnessed by the universal popularity enjoyed by the argument that aims to defend the practices of exploitation of non-human animals in the name of the natural order and the “place” of the human species, in the food chain in particular:
(...) The entire human food chain is based on what some call “speciesism”. Food chains in nature demonstrate that each species lives at the expense of one or more others. The use of animals by humans is natural, just as in other food chains34.
Life is a uniquely complex phenomenon. We assume that the fundamental physical laws on which it depends are known35. But there is a gulf between the level of elementary particles and that of an organ like the eye. To really explain a complex phenomenon is to make it intelligible; it is not enough to demonstrate, let alone postulate, that it could be recreated by a gigantic simulation based on the laws of particle physics. The same is true when we start out with genes rather than particles. To state the elementary rules governing genes, to note that some will be more reproduced than others and therefore will endure while others disappear, and that this necessarily explains the eye since the eye exists, is not to explain the eye. A series of intermediate levels of understanding are missing.
Dawkins is aware of this need to avoid what he calls “precipice reductionism”36. Much of his work is an attempt to construct intermediate levels of explanation to make life intelligible. I think his results fall far short of this ambition; but this only means that the fruitfulness of Darwinism is ahead of us, not behind us. Dawkins, like others, but less than others, tends to fill the gaps by the clandestine, almost spontaneous and inevitable reintroduction of the same finalism that has served us for millennia as the sole explanatory principle. At least he makes the effort, inconstant and often awkward, to distinguish the finalist metaphor from reality; whereas even today most of his colleagues have no problem asserting that they have explained a phenomenon as soon as some “utility” has been found for the individual or the species, when it is not for the entire ecosystem. On the other hand, even when Dawkins makes organisms into “survival machines” refined since the dawn of time by primitive replicators, he refrains from deducing from this ontology ethical conclusions. The subjectivity that he allows himself to attribute in one way or another to various biological objects does not replace and render invisible the real subjectivity of sentient beings, which alone can serve as a basis for ethics. I think that ethics wins, and so does biology, which has every interest in accepting that Darwinism today constitutes more than just an achievement: it remains a challenge, that of building a real understanding of life, this astonishing phenomenon.
1. I am less familiar with the ideas of Edward O. Wilson, but it appears that the two authors should not be confused.
2. The Selfish Gene, Oxford University Press, Oxford, 1976, 1978. I will refer to this book as SG. I will also often refer to another book by Dawkins, The Extended Phenotype, Oxford University Press, Oxford, 1982, 1999, under the abbreviation EP.
3. Mary Midgley, “Gene-juggling”, Philosophy 54 (1979), pp. 439-458, quote from p. 439.
4. “In Defence of Selfish Genes”, Philosophy 56 (1981), p. 557; Dawkins quotes here from SG on page 4. The italics and the break in brackets are Dawkins'.
5. Dawkins clearly considers behaviour as a non-subjective notion; this point could however be subject to long discussions. It is true that in physics we speak of the behaviour of objects without subjectivity; the accelerated downward movement is the “behaviour” of a stone in a field of gravity. But it is doubtful that Dawkins would consider the fall of a lion who inadvertently passes over a cliff as an example of altruistic behaviour – although by behaving in this way (by falling rather than levitating) the lion decreases her own welfare (chances of survival) while increasing that of the gazelles! This is because such a fall is perceived as “passive”, whereas the notion of behaviour is in common usage linked to that of activity. However, the active/passive opposition only pushes the problem back: for how could an object devoid of subjectivity be anything but passive? Note also Dawkins' confusion when he attempts to distinguish between “active” and “passive” replicators (EP, pp. 83-84) – he himself seems to admit that this distinction does not work, but retains it nonetheless.
6. The term survival is in fact itself implicitly redefined as simply continuing to exist – for genes do not in fact live, and so cannot in any meaningful sense survive.
7. SG p. 20.
8. SG pp. 210-211.
9. “Constraints on Perfection”, EP, pp. 30-54.
10. Richard Lewontin's definition quoted by Dawkins in EP, p. 30.
11. “Environment” should be understood here as including the rest of the genome.
12. Dawkins no longer speaks of genes, but simply of DNA, when it comes to untranslated genetic material. This distinction seems to have little basis in his own definition of the gene, which is independent of its translation into protein molecules (cf. SG p. 28 ff., or EP, p. 81). I keep his terms, however, for convenience.
13. SG p. 47.
14. Genesis 1:24 (King James version).
15. SG, p. 21.
16. “Sociobiology: the New Storm in a Teacup” in S. Rose and L Appignanesi (eds.), Science and Beyond, ed. Blackwell, Oxford, 1986, p. 66, retranslated here from French. See also “In Defence of Selfish Genes”, op. cit. p. 572, where he states that he understands this characterisation of ourselves as “robots blindly programmed to preserve [.... ] genes” in a literal sense (albeit redefined).
17. Dawkins appears to dispute that purpose implies subjectivity (cf. GE pp. 53 ff). Such a position seems to me untenable.
18. SG, p. 37.
19. SG, pp. 29 ff; see also EP, Chapter 5.
* In the original, 2001, text of this article, I quoted Dawkins from the published French version (Le gène égoïste, Odile Jacob, Paris, 1996, translation by Laura Ovion), without checking on the original text. Unfortunately, this version has two serious errors: “That is their business” is translated as “c'est leur rôle” (“that is their role”), and “denizens” as “citoyens” (“citizens”). Despite this, I prefer to translate this paragraph from French without corrections, even if the words I cite are not Dawkins' and the collective finalism that I criticize was, in fact, introduced by the translator.
One can, however, wonder why so many people tend to read Dawkins as the translator did, and view him as attributing a collective will to the genes.
20. “La nature ne choisit pas” (“Nature does not Choose”), Cahiers antispécistes No. 14, 1996, reproduced in this volume on page 89. Dawkins himself reports (EP, pp. 179-180) what he calls a “fascinating” (indeed!) excerpt from a letter from Alfred Wallace to Darwin (1866) which carries essentially the same criticism against the term “natural selection”. Darwin partly agreed, but thought that only “a few minds”, increasingly rare, would be led astray by this finalism. What is the situation today?
21. EP, p. 81.
22. SG p. 48.
23. “Gene-juggling”, op. cit., quote from pp. 452-453 (words highlighted by Midgley).
24. Stephen J. Gould, quoted by Dawkins in “Sociobiology: the New Storm in a Teacup”, op. cit. p. 66, retranslated here from French.
25. Jean-Jacques Kupiec and Pierre Sonigo, Ni Dieu ni gène. Pour une autre théorie de l'hérédité, Seuil, Paris, 2000, my translation to English.
26. Op. cit. pp. 129, 150 and 197. These three extracts are by Pierre Sonigo.
27. See for example Prisoner's dilemma on Wikipedia.
28. See for example Inclusive fitness on Wikipedia.
29. EP, p. 51.
30. See SG, pp. 154 ff.
31. The central command of Christianity is: “care for your salvation”. For a certain liberalism, the general interest is best served when each economic actor behaves in a strictly selfish manner. Communist society is supposed in the Marxist project to allow the contradiction between general and individual interest to be “overcome”; here again altruism has no place, and already now Marxists declare any altruistic behaviour suspect, as necessarily a mask for less noble motivations. The result of these attitudes is a form of banishment of altruism. Let us note some examples. In France, once a company is declared to be “profit-making”, managers who dare to take into account considerations other than profit maximisation would be committing a “misuse of corporate assets”; capitalist companies are therefore ruthless by legal obligation. According to Gould in the quoted passage, natural selection dictates that organisms act in their own interest; selfishness is thus presented as a prescription. And the human being who altruistically refuses to eat non-humans is often accused of immorality because such behaviour is “unnatural”, contrary to the place of humans in the food chain and might lead to disaster if it became widespread.
32. See for example Dawkins' clarification in “Sociobiology: the New Storm in a Teacup”, op. cit. pp. 76-78.
33. I have in mind certain articles in the Figaro Magazine of the late 1970s, if my memory serves me well, where the thesis of “useless” DNA was presented as the final proof of the selfishness of genes, and therefore of nature, and therefore of the human being.
34. Excerpt from “What are animal rights and what are they not?”, website of the pro-speciesist association Brok, http://www.brok.be/ This kind of argument abounds in any good humanist book.
See also: “The Subjective Is Objective: Taking Sentience Seriously” (long), or “Science and the Denial of Animal Consciousness: Mind-Matter for Animals Matters” (shorter).
35. I tend to think that there are essential aspects of the laws of physics that have yet to be discovered, particularly in order to explain the phenomenon of subjectivity. For an argument in this sense, see Roger Penrose, Shadows of the Mind. A Search for the Missing Science of Consciousness, Oxford University Press, 1994.
36. See “Sociobiology: the New Storm in a Teacup”, op. cit. pp. 72-75.