The distinction between ”gene” in the sense of an allele at some locus and ”gene” in the sense of a dna sequence with a name and some function seems easy enough, but still causes a lot of confusion, both in popular and scientific literature.
This was very clear a few months ago when science journalist David Dobbs published his ”Die selfish gene, die” and a few weeks of debate broke out. In my opinion it’s not a particularly good piece, but I agree with Dobbs that the ”selfish gene” metaphor sometimes invites misunderstandings. The article itself displays a few of them, when it suggests that evolution and genetics as understood before the age of microarrays are somehow at odds with the importance of gene regulation or phenotypic plasticity. I suspect that many of these problems stem from the double meaning of the word gene. Other examples are found in headlines claiming that researchers have found the gene for something or the confusion about the word pleiotropy (Paaby & Rockman 2012).
When Dawkins wrote about the selfish gene, he did not mean the selfish dna sequence encoding a protein; he meant the selfish genetic variant causing differences in fitness between individuals. (Or rather, a set of genetic variants in sufficiently close linkage to seldom be separated by recombination.) The book is not about molecular genes. As anyone who actually read it knows, it deals mostly with behaviour using game theory approaches. This does not mean that Dawkins denied that there are actual molecular genes doing the mechanistic work, but that he analysed the situation mostly on a different level. And had he chosen to write only about known sequence variants with adaptive effects on behaviour it would have been a very short book.
Of course the word ”selfish”, while I agree that it is the proper word in the sense that Dawkins intended, is great for those who want to point to instances where people are horrible to each other and tell you that it’s all because of evolution. But I think that is a bigger issue that will not be solved by tweaking popular science metaphors. By the way, that is completely contrary to Dawkins’ intentions, which were to popularise the evolutionary models that explain why animals are not always horrible to each other, even though their behaviour is shaped by natural selection.