Q. An article in the Washington Post (“On the Web, Research Work Proves Ephemeral,” by Rick Weiss, November 24, 2003) reports that URLs often become obsolete. It also says that there are lots of errors even in citations to conventional sources. What’s a researcher to do?
A. It is important to remember that the typical URL (uniform, or universal, resource locator) is essentially an address—something less than a permanent identifier for a source—and, like the Dewey decimal classification number on a book that sits on a library shelf, its primary function is to tell you where an item is, not what it is. A URL should never be the sole component in a citation: always include information such as author, title, and publication date to the extent these can be determined. The Post article cited above, for example, will continue to be findable from its title, author, and publication date—through a library database or a search engine—long after a particular URL for the source is dead. For less formally published sources, save or print out a copy of the source for your files. As for errors, authors and researchers should transcribe source information and include all the required elements as if there will be no opportunity to check for accuracy later. And if computers make some information ephemeral, they also help ensure that, for example, a typo in the volume number of a journal won’t provide an overwhelming obstacle to finding the article from the rest of the citation—or to checking the accuracy of the citation at manuscript stage. (See chapters 14 and 15 in CMOS for a full discussion of documentation, including information about how and when to incorporate URLs into citations.)