Just because your papers no longer rack up citations doesn't mean people aren't reading them--at least for some sciences. That's one implication of a look at access data by JSTOR, an online archive of 117 journals mainly in the arts and social sciences (www.jstor.org). The informal review of views back to 1997 found that citations don't always correspond to usage. For example, the fourth most accessed paper in economics--a 1973 article in the Journal of Political Economy--was cited only four times between '97 and '99 (and 14 times since 1974), compared to scores of citations for other, less read papers. That suggests that some papers that aren't pushing a discipline forward may nevertheless be very valuable for teaching, writes JSTOR's Kevin Guthrie in a recent conference paper. He concludes that "citations do not provide anything like a complete picture of the potential usefulness of a journal article."
Other old papers are also being widely read, especially in math, where the most viewed papers are on average 32 years old. That's no surprise to mathematicians, who note that theories don't go out of style, say JSTOR staffers. But they say the results might turn out differently for "hard" sciences, where fields move more quickly.