Irv Sher and I when we did the genetics project we created the initial... we had the journal list of really important biomedical journals from Current Contents. We couldn't have published Current Contents. At that point it was, let's say, 4 or 5 years old. Medical librarians, Charles [unclear] all those people said these are the 100 most important journals that we were covering, all along, okay. Now, would there have been a really significant journal that they would have said, don't put it in there. So, it's kind of almost silly that they would think was such a huge exercise to get the most important journal. We all knew what the journals were, whether we were called subjective or not. Now Estelle Brodman who was a medical librarian and I loved her as a person and as a historian, had this, she didn't like the numbers game either. She thought that, she did a paper on choosing physiology journals and she objected to the fact that our ranking gave too much weight to certain types of journals, I guess review journals because of the, you know, their inordinate citation. Well, they're still important. Beyond that, anybody knew what the leading hundred... I guess once you get past the first few hundred titles then you can get into these squabbles about what is or isn't the next journal to add. So, Irv and I said, what are we going to use as the criteria. We know what we start with. And then we did this journal citation analysis, the primitive JCR and we called it, it was the Journal Citation Index when we first did it and we did it because we thought it would better to maybe publish the citation index, not sorted by author but sorted by journal. Which, in a sense, was the way Shepard's Citations was organised, you know. They put more emphasis on the vehicle rather than on the individual, the individual case or whatever. So, you'd have it by journal. Well you could of had that. So, we did a journal sort in order to see what it would look like. There were a number of advantages to the journal sort if we had done it. We could have eliminated... at that point, we recognised that if an author had his name spelt differently, when you did a journal type sort, it would be brought together regardless of the spelling of the cited author, okay. But, we had to say well, most people were used to citing author journals, that's why we went that way. And then it was just a trying to make a decision between these small and big journals so, this was a way to normalise size, okay. Now, we did think about the concept of impact, even in 1955, we mentioned that in our paper but it wasn't the formal impact factor. Today when people talk about impact factor, they don't always mean journals, they're talking about an author's impact, which is a very confusing thing to do.

[Q] Or institutional impact or country impact, all sorts of things.

And no matter how you slice it, there's always going to be somebody who's got an objection. All along there was always the criticism that it was mostly English that we're biased the game in favor of the English language. Then, whenever you did a study to disprove that, they'd find some other objection. When it wasn't, you should be comparing European journals and now the Germans object when we evaluate their journals. Then you tell them, if you were a German scientist, where do you think you're going to publish a really hot paper? In the German journals or in the international journals? It's the authors that really determine the game, not us. So we are selecting where the action is so to speak. In the beginning, we were emphasizing the fast moving fields, molecular biology, biochemistry and so forth and so there was somewhat that bias. So, what we maybe didn't do from the early days is quickly enough pick up on the slower moving fields. I mean to this day, I question whether we have too many, let's say, taxonomy journals missing from the ISI coverage but there may be some more fields like geography, but how you would pick up that stuff I don't know, there's so much low impact material that after a certain point, you can't even discriminate. You know what's the difference between something with an impact of point one or point two? Yet people will make a big thing out of that.