The Selfless Gene

The Selfless Gene

The logic of selfish gene theory is impeccable, but its usual formulation leaves something to be desired in point of strict and literal accuracy. I propose to remedy that. No harm will be done to the theory thereby; we will merely clarify its exact import. First, some terminology: by “gene” I will mean a particular piece of matter that exists on an individual strand of DNA—a specific localized chemical lump. I will not mean the type of gene of which this is a particular instance: this is a general form that can be instantiated in many individual instances; it might be viewed as the information contained in the chemical instance or the general type of which there are many instances (my genes, your genes, etc.—construed as local particular lumps of matter in individual bodies). We might call these gene tokens as opposed to gene types. This distinction is vital to the point I want to make, because we might mean the token gene is selfish or the type gene is selfish. Is the gene token (that particular lump) selfish?

To answer this let’s consider an ordinary case of selfishness: a person who cares only for himself and no one else—healways comes first. Such a person might coexist with copies of himself—replicas, twins. Perhaps they have been created by cloning him; they are exactly like him, but they are not him (that individual). Let’s suppose he doesn’t give a damn about these copies; he cares only for himself. He would happily let them die if it suited his own selfish purposes. Now contrast this selfish person with another person who also has many copies, but he does care about them; he goes out of his way to help them out, share with them, even sacrifice himself for them. This person is selfless not selfish; he is an altruist not an egoist. But suppose he is indifferent to anyone who is not a copy of himself; he is selectively selfless. It would be wrong to call him selfish, but his selflessness is sharply limited to his doubles, duplicates, twins. His altruism is confined to a certain select group of individuals: these he promotes and supports, feeds and protects—though anyone else gets the cold shoulder. It is this individual who best represents the concept of the selfish gene not the outright selfish person. The gene token seeks the welfare and survival of its copies not itself alone; it will go when the body it is in disintegrates, and it knows it. But its replicas will live on—and that is its mission in life. It may even sacrifice itself for the sake of its copies—as when a mother gives food to its babies even at a considerable cost to herself. It produces as many copies as possible, so that its type survives, possibly immortally; but it itself will perish with the bodily vehicle in which it exists. The copies will exist even when the individual gene that made them has gone. This gene is not selfish in any reasonable sense—though we might say the gene type is selfish because it sees to it that this type will survive into the future.

We really have two questions: are gene tokens selfish and are gene types selfish. No to the former, and yes to the latter. It depends on what we are referring to. I rather think people generally mean the former, in which case they misspeak, because the tokens are selfless not selfish. The token is selfless in the way the individual animal is selfless: the mother is concerned to protect her offspring copies even when this goes against her own individual interests (she might die as they suckle at her teats). It would be strange to say she was selfish because her type is trying to protect that type; she is protecting her copies not herself, as opposed to being a type that is protecting itself. So, the so-called selfish gene is not really selfish but selfless—altruistic, other-interested. Our genes (individual tokens) are actually generous and other-directed, but only within the class of recipients limited to type-identity with them. They are like a person only altruistic within his tribe or family, only more strictly. The Selfish Gene could have been called The Selfless Gene and said essentially all the same things, reserving for a footnote the observation that gene types can be said to be selfish in an extended sense of “selfish” (we don’t say person types are selfish but only person tokens). Still less is the information inside a person itself selfish. It is really a lot simpler simply to say that genes (those bits of matter) are selfless, adding that their sole focus of altruistic interest is their own copies. They don’t survive bodily destruction, though their replicas do. They are not physically passed on, though physical entities just like them survive in future bodies. It is certainly not true that my token set of genes is literally made up of the same material entities that existed in my ancestors; what survives are just the old types, or the information they contain. As I say, this doesn’t detract from the correctness of selfish gene theory, though it does recommend changing the rhetoric slightly. And I think the book would have attracted a lot less misunderstanding if it had been called The Unselfish but Selective Gene.[1]

[1] For philosophers steeped in the type-token distinction in application to the mind-body problem this paper will be pretty straightforward, but that distinction is not always apparent to people less prone to logic-chopping. I have tried to make it as clear as possible, based on past experience. The logical point is simply that “x is a replica of y” doesn’t entail “x is (numerically) identical to y”—in fact the opposite is true.