Some Problems Related to Corrections of Error in the Scholarly Literature

The topic of correction of scholarly error in academia is a vast and complicated one. This short article is an introductory discussion, or brief summary, for a more elaborate study that is underway.According to the rhetoric of the academic establishment, there is really no serious problem involving correction of scholarly error. The often repeated phrase "Science is self-correcting" implies that there is no problem for science at all, since there is an automatic process that assures-or at least strives for- quality control. With peer review in place, errors are detected by reviewers/referees, and by editors, and are therefore not published. If by chance some errors happen to slip through the peer review quality control safety net, specialists in the specific academic discipline will spot the errors and corrections will be made.But despite the alleged self-correcting process of science, and the quality control afforded by peer review, much erroneous material is published in the scholarly literature, and some errors persist for long periods of time. For instance, in the case of childbirth fever, Semmelweiss is often given credit as the scholar who discovered the cause and prevention, and who therefore corrected the specific serious medical error that was causing so much harm. Yet, other medical researchers before Semmelweiss including Charles White, Oliver Wendell Holmes (relative of the famous jurist), and Alexander Gordon, came to very similar conclusions. In this case, de Grazia (1984) observes that "It took about a century from White's obsessive insistence on cleanliness in Manchester's lying-in ward to consensus about a matter that should have been simple enough to grasp" (p. 29).A very interesting study relating to correction of error is Serge Lang's 1981 book, The File, which is subtitled Case Study in Correction. Lang detected defects and errors in a specific sociological survey of the academic profession. What began as an exchange between Lang and one of the authors evolved into a controversy-and a book of about 700 pages-that "involved a good part of the education network" (p. 1). Had corrections been made at the beginning, this book would not have been written at all. In fact, according to Lang, the book resulted from the "combination of an attempted correction with the refusal to make the correction..." (p. 1). He also notes: "I am bothered by the ... obstructions which prevent correct information from being disseminated. These obstructions come about in many ways-personal, institutional, through selfimposed inhibitions, through external inhibitions, through outright dishonesty, through incompetence-the list is a long one" (p. 2).Such obstructions to correction of an error can be grouped into several categories: 1) lack of volition or lack of expertise, on the part of specialists, to make corrections; 2) paradigm dependence, paradigm protection; 3) conflict of interest; 4) advocacy research and ideology; 5) disdain, on the part of establishment scholars, including editors and peer reviewers, for scholars who attempt to publish corrections. (Here, mention and discussion of these categories will be very brief. There are some overlapping aspects for these categories.)Lack of Volition, Lack of ExpertiseThe infamous so-called Baltimore controversy is an interesting example. In effect, a junior researcher tried to publish corrections of error that she had detected. There was resistance to such corrections. The situation escalated and dragged on for years, with questions of correction of error giving way to questions of scientific misconduct. Investigations were made at Tufts University and at MIT and eventually in the United States Congress and at various levels of United States government agencies. A governmental Appeals Board overturned prior findings of misconduct, but at the same time stated that the article in question was "rife with errors," including errors that "despite all these years and layers of review, have never previously been pointed out or corrected" (Science, 1996, p. …