AuthorsD. I. K. Sjøberg
TitleConfronting the Myth of Rapid Obsolescence in Computing Research
AfilliationSoftware Engineering, Software Engineering
StatusPublished
Publication TypeJournal Article
Year of Publication2009
JournalCommunications of the ACM
Volume53
Issue9
Number9
Pagination62-67
PublisherACM
Abstract

By 1999, Microsoft had already become the world's largest company measured by stock-market value. Now, the Internet is offering us new services every day. Computing technologies have moved on and are rapidly changing our social, political, economic, and cultural worlds [13]. Accompanying this observation is the common belief that research in computing makes progress more rapidly, and just as rapidly becomes more obsolete, than research in other disciplines. One indicator of how fast research becomes obsolete is how quickly reports of it cease to be cited in the literature. Common measures of this phenomenon are cited half-life, citing half-life, and the Price Index. Interestingly, and contrary to popular belief, these measures indicate that research in computing does not cease to be cited any more quickly than research in other disciplines. Therefore, research in computing does not become obsolete any more quickly than research in other disciplines. The extent to which this is the case is important for a number of reasons: 1) Although there have been tremendous achievements in the history of computing, there is an exponentially increasing demand for complex, reliable, robust, and usable hardware and software systems in society. The advances in computing technology that need to be made to meet this demand depend on the long-term funding of fundamental research [11]. However, it seems very difficult to convince funding bodies to support long-term fundamental research programmes and projects in computing. One reason may be the already rapid development in computing applications. This may be taken to indicate that research in computing is less difficult than in other disciplines and that as a result, progress can be made with less funding that other disciplines receive. Hence, as has been reported in the context of NSF, when fighting for research money, we computer scientists must argue on the one hand for the practical conclusion that society has a strong need for the results of our research, while on the other, for the identity of computer science as a basic research discipline to maintain respect within the scientific community [18]. In particular, in a university setting, when competing for funding with researchers from other sciences, we must counter the argument that basic research funding in computing will not be prioritized because everything that is useful is being done both faster and better by the IT industry anyway. 2) Literature that is still relevant may be considered obsolete, and thereby ignored, merely on the grounds that it is old. As a researcher and journal editor, I have found that reviewers comment on “old references” fairly frequently, and as a supervisor, I have found that PhD students often are reluctant to read older literature.

DOI10.1145/1810891.1810911
Citation KeySimula.SE.435