The Extraordinary Growing Impact Of The History Of Science

Old scientific papers never die, they just fade away. Or they used to. Now electronic publishing has made old papers as easy to find as new ones, and the effect on science is profound

Walk through the archives of any academic library and you will find dusty shelves of scientific journals dating back to the early 20th century and sometimes beyond. It’s hard to know the last time many of these volumes were opened or whether the information they contain has long been forgotten.

That’s in stark contrast to the way most scientists access scientific papers today; in a matter of seconds after a straightforward web search. Never has it been so easy to look up a circuit diagram, learn about gene therapy or read the latest papers about black holes.

That’s why many scientific publishers have digitised their archives to make the entire history of their publications available online and as easy to search as modern papers.

That raises an interesting question — if old papers are now as easy to find as modern ones, are they having as great an impact?

Today we get an answer of sorts thanks to the work of Alex Verstak and pals at Google. These guys have studied how often older articles are cited in modern papers and how this has changed since the advent of electronic publishing in the 1990s. Their conclusion is that older papers are having an increasingly important impact on modern science — that the distinction between old and new, between the historical and the modern, no longer creates a division in science.

These guys base their work on a database of citations in scientific papers published between 1990 and 2013 in 9 broad areas of research subdivided into 261 subject areas. For each discipline, they then plotted the percentage of citations to papers that were at least ten years old.

The results show a clear trend. “Our analysis indicates that, in 2013, 36% of citations were to articles that are at least 10 years old and that this fraction has grown 28% since 1990,” say Verstak and co. What’s more, the increase in the last ten years is twice as big as in the previous ten years, so the trend appears to be accelerating.

The results solve an ongoing conundrum among researchers involved in scientometrics, the study of science and scientific research. Some of these researchers have long argued that the ongoing digitisation of historical papers should automatically ensure that they are cited more often. Others point out that there has been a huge increase in the number scientific papers published in recent years so historical papers should be a smaller proportion of the total and therefore cited less.

The work of Verstak and co shows that the former effect has won out. “Now that finding and reading relevant older articles is about as easy as finding and reading recently published articles, significant advances aren’t getting lost on the shelves and are influencing work worldwide for years after,” they say.

There are one or two interesting wrinkles in the data. These trends appeared in 231 out of 261 subject areas. But many of the subject areas that experienced a decline in older citations were part of two broader areas: chemical and materials sciences, and engineering.

Consequently, these broad disciplines show almost no increase in old citations. Just why citation trends should differ in these disciplines isn’t clear.

Nevertheless, that’s a fascinating result. Instead of fading into the past, the history of science is becoming a growing influence on current practice. That’s clearly no bad thing.

But here’s the thing — if the history of science can have a bigger impact on the present, why not history in general? Could it be that this work is the first sign that the digital age is likely to make history a much more important part of everybody’s present because it will be just as easy to access electronically as the recent past? Answers in the comments section please.

Ref: arxiv.org/abs/1411.0275 : On the Shoulders of Giants: The Growing Impact of Older Articles