I have a confession. I like to read science fiction. Oh, I also read literature, and mysteries, and historical fiction, the works of Borges and Calvino, and poetry (although I don't always get poetry, except for Charlie Bukowski), and lots of other things. But I like science fiction too.

When I was just a kid, I read everything by Ray Bradbury, but didn't realize that he and Edward Banks were one and the same. I wasn't surprised to learn that the brilliant polymath Isaac Asimov published David Starr, Space Ranger under a pseudonym (Paul French), although it surprised me a great deal when I learned in 1978 that James Tiptree, Jr, mas macho author of Your Haploid Heart and a wealth of other stories, was actually ex-CIA agent Alice Bradley Sheldon. Go figure.

Why am I mentioning all this? Well, I've been thinking about authors. Pseudonyms are fun, and I, for one, see nothing wrong with them. They can even be useful, as in the case of Alice Sheldon, who penetrated the male dominated world of SF in the 1950s. (In case you were wondering, `Mole' is not a pseudonym. My parents simply lacked imagination.)

In biomedical research, when we publish something, especially a primary paper (one that actually contains data), we often have many authors. And we don't use pseudonyms, in general or at least I don't think so (I may be in for a few surprises – there is a rumor that Louis Pasteur was actually Victor Hugo, but I don't give it much credit). Pseudonyms may not be a bad idea, though; in this era of PubMed and impact factors, a well-chosen pseudonym could prevent awkward duplications of surnames and initials and facilitate monitoring. One prominent molecular biologist I know was confused to discover that he was a highly cited physicist. But no, I can't recommend pseudonyms.

But that isn't what I wanted to talk about. I wanted to talk about who, exactly, should be an author on a paper. I do know of two fictitious authors of biomedical papers: one was Z. Zebra, added to the end of a rather long list of authors, and another was Galadriel Mirkwood, who was a dog (hence not really fictitious I suppose), but these were in less regulated times. But mostly, I suspect, authors on a paper represent real people.

It has always been the case that co-authors can suddenly appear on a draft of a paper, often to the surprise of the first author who actually did all of the work. This can be for readily justifiable reasons – or not. It's the `not' that I want to talk about.

It generally goes like this. We're putting together the manuscript, and realize that we had obtained a quantity of reagent from another lab with the understanding that we would include the provider among the authors. It seemed fair – we couldn't do the work without the reagent, and it would take too much time and effort to make it ourselves. Upon hearing back from this co-author, we are informed that we must include two more co-authors, who all participated in generating the reagent. We add them, and then one of them wants to include three more people. One is his dog.

Then Joe and Brunhilda from the lab next door point out that they helped to design one of the experiments over cocktails one Friday night, and Alan the bartender should also be included, since the drinks contributed significantly as well. And, of course, we also have to include J&B's boss and probably Alan's as well.

This is not new, nor is it a problem that some have not tried to address. Some journals require that each author's specific contributions be included. But of course, this is easily circumvented by simply making things up (Alan provided discussion and valuable reagents, for example, even if the reagents were particularly good martinis and the discussion was `nice night').

So why do we put up with this? Well, first of all, we know it doesn't matter much to the first or last author, who will receive due credit. A simple solution, and one that I have practiced, is to leave the decision up to these two individuals – if they don't mind, then it is probably okay. (In general, the way I apply this is, if someone wants to include me on their paper for my contributions, if both the first and last author feel that I've genuinely helped, then I'm okay with it.) Does it really not matter, though? It's not such an easy question. While it doesn't dilute the impact for the authors who are first or last, it can have that effect on middle authors who really did do a lot of work. They should get more credit than they are getting, and that credit is diluted by this practice.

But, for many of us, it is much easier to add authors than to risk displeasing one, especially if that one has or is perceived to have significant clout. And it leaves a bad taste. I have a friend with such clout who once informed me that he always demands authorship in exchange for reagents, even if he simply buys them (he contends that it's the other lab's fault for not taking the time to look up where they can be purchased). I'm not entirely comfortable with this, especially as the ensuing publications contribute, a lot, to his significant clout.

So what can we do? Of course I have an answer. I call it the `Rule of Two.' I didn't make it up; it's something I learned from my wise friend Professor Rabbit. For someone to be an author on one of our papers, they have to convince us that they have contributed to the work in at least two ways. These can be any two of the following: technical work, experimental design, experimental results, valuable reagents, writing the paper, and insightful discussion. (Note that `discussion' doesn't cut it; they have to have contributed an illuminating idea to the work.) This Rule of Two works pretty well when Joe and Brunhilda point out that they helped to design some of the experiments; we then can say, “Okay, that's one.”

But Mole, you say, that doesn't help me when I'm effectively blackmailed into providing authorship in return for one thing that I actually need. If I come back and say, “that's one” they may simply refuse to give me the valuable thing. And if I subsequently renege on the agreement, however coerced, I may be in for trouble. At the very least, I won't get help from them again.

It's true. But, again, I have a solution – one that I'm planning to use myself, which is why I'm thinking about authors. Recently, I agreed to include two scientists as authors on any papers that I write that use a genetically modified animal I obtained from them, despite the fact that they had published on this animal in journals that require them to share these animals (they did share them, provided that I always include them as authors). So here's what I'm going to do. After their names in the author list I will place a footnote that states simply, “These authors are included by agreement for providing the animal used in these studies.” Perhaps they won't care, but it will make me, and my real co-authors, feel a bit better, and no one who reads the paper will think that these two contributed in any other way. So when we are required by agreement to include an author, we can simply inform the reader of precisely why they are authors. Perhaps we should do that for all the authors. Indeed, some journals do – maybe they all should. I can't see anything wrong with that – unless, of course, scientists who contribute reagents start to dictate how they should be acknowledged for their contributions.

But that's science fiction. Or is it...?