Papyrus to PowerPoint (P 2 P): metamorphosis of scientific communication.

researchers used to visit libraries. More than half of online sessions end with a full text download, which is likely to look exactly like the article would in the paper journal. Users say they are interested in the added functionality of the web version, but even following links—which comes out top in users’ assessment of features—is a distinctly minority activity. Once users have downloaded what they want, they apparently annotate it and file it in filing cabinets and folders just as they always did. If this remains the limit of what users do, it makes the whole online publishing project look like a not very interesting storm in a teacup. Derk Haank of Elsevier Science, which publishes more than 1000 online journals, apparently concurs. While emphasising the need for more content and linking, he says that “no more features” are needed. And once upon a time IBM estimated the total worldwide market for computers at half a dozen. This is a familiar pattern: when confronted by a new innovation people tend to overestimate its short term effects while underestimating its long term effects (fig 2). With the benefit of hindsight we can surmise that our published predictions of five years ago fell into the early phase—when hype outstripped reality. If the pattern holds true, we should expect reality ultimately to turn the tables.