The Internet has made it far easier for professionals to access and search the scientific literature. Unfortunately, it has also increased the chances that we will lose part or all of that literature.
When I was young, I imagined that everything I wrote would be preserved forever. Future biographers would seek out every letter, diary and memorandum to capture the essence of my creativity. My first laboratory notebook still captured the same emotions. On page one I had printed, very legibly, the following preface: "To Posterity: This volume contains the authentic record of ingenious and original chemical research conducted by Robert Shapiro, currently a graduate student of organic chemistry at Harvard University."
Reality gradually whittled down my grandiosity, and I recognized that my published papers had the best chance of survival. The New York University library carried bound journals that dated from the 19th century, and the articles thay contained had obviously outlived their authors. As the number of my own published works grew, curiosity chose me to select one of them and track its impact. I deliberately picked one of minor importance.
A generation ago, a persistant PhD student and I had failed in an effort to synthesize a new substance of theoretical interest. We had however prepared some other new compounds and improved some methods, so I wrote a paper that was published in 1969 in The Journal of Organic Chemisty. Had our results ever mattered to anyone? Using new computer-driven search tools, I could quickly check whether it had had ever been noticed. To my surprise, I found that 11 papers and some patents had cited our publication, up to 2002. In one instance, our work provided a starting point for the preparation of new tranquilizers. I imagined that in the distant future, other workers might pull the appropriate volume off a library shelf and find my work to be some help. I did not forsee that such bound volumes might no longer exist.
The Journal of Organic Chemistry started in 1936, and continues up to the present. Its demands on library shelf space have increased over time: the first volume contained only 583 pages, while the 2009 edition had 9680. The arrival of the Internet rescued libraries from the space crisis created by the proliferation of new journals and the vast increase in the size of existing ones. Many paper subscriptions were replaced by electronic ones, and past holdings were converted to digital form. It is not hard to imagine a future time when paper copies of the scientific literature will no longer exist. Many new journals are appearing only in digital form.
This conversion has produced many benefits for readers. In the past I had to leave my office, ride an elevator, walk several blocks, take another elevator, and make my way through a maze of shelves to find a paper that I needed. Occasionally, the issue I wanted was being used by someone else or had been misplaced, and I had traveled in vain. Now I can bring most papers that I want onto a computer screen in my office or at home in a matter of minutes. I can store the publication in my computer, or print out a copy if I wish. But with this gain in the accessibility of the literature of science has come an increase in its vulnerability.
Materials that exist in one or a few copies are inherently at greater risk than those that are widely distributed. A Picasso painting might be destroyed but the Bible will survive. Alexander Stille in The Future of the Pastreported that the works of Homer and Virgil survived from antiquity because their great popularity lead them to be copied and recopied. On the other hand, only 9 of Sophocles 120 plays have survived. Before the Internet came into play, I could take pride that my each of my papers was present in hundreds or thousands of libraries across the globe. Its survival into the future was enhanced by the protection afforded by multiple copies. The same applies, of course to the remainder of the scientific literature.
Thousands of paper copies of the literature have now been replaced by a few electronic records stored in computers. Furthermore, the storage medium is fragile. Some paper manuscripts have survived for centuries. The lifetimes of the various discs, drives and tapes currently used for digital storage are unknown, but are commonly estimated in decades. In some cases, works available only in electronic form have disappeared much more rapidly for another reason — lack of maintenance of the sites. One survey found that 12% of the Internet addresses cited in three prestigious medical and scientific journals were extinct two years after publication.
Such difficulties are unlikely to affect prestigious sources such as the Journal of Organic Chemistry. But material stored only on the Internet is far more vulnerable to destruction than the same material present in multiple paper copies. Electrical breakdown can disrupt access for a time, while cyberterrorism, civic disturbances, war and a variety of natural catastrophes could destroy part ar all of the storage system, leading to the irretrievable loss of sections of the scientific literature. Anton Zeilinger wrote in a previous edition of this series that a nuclear explosion outside the earth's atmosphere would cause all computers, and ultimately society, to break down.
How has this changed my thinking? I no longer write with the expectation of immortality in print. I am much more tempted to contribute to Internet discussion forums, blogs, and media which may not persist. I seek my reward from the immediate response that my efforts may bring, with little thought to the possibility that some stranger may see my words centuries from now, and wonder about the life that was led by the person who wrote them.