Lucas Estevam points us to this interesting article by Keith Collins.
The article discusses several interesting features about how/why retracted papers may get cited – focusing on the behavior of authors (legitimate and less legitimate reasons for citing retracted works) and journals (failure to make retractions evident) – but it omits perhaps the most interesting aspect of the problem. This is the role of reviewers. I think it illustrates the failure of the refereeing process that so many published works cite retracted work for less than legitimate reasons.
Journals don’t exactly make it easy to spot retracted papers.
Ok, who else read “retarded” at first glance?
There really ought to be at least two kinds of citations:
1) The usual positive kind.
2) A negative kind, saying that one is critical of a source, or it has been retracted, or both.
Perhaps noted visually via the hyperlinked reference in the paper: blue is good, red is bad? That would be kinda funny.
Although maybe we should make this both more legible and more nuanced: “blue-light cyan-medium gray-red, in that order”
Current citation-system is obsolete and dysfunctional. It’s an academic relic of past centuries and a former very small world of information and authors. The volume of information has expanded far beyond its capability and design.
What is the fundamental purpose of a citation-system ?? Who benefits from the relic still in use?
Any communication has Transmitters and Receivers. Primary purpose of current citation-system is to ensure authors(transmitters) get credit for their words/work; in a small information world, the personal reputation/integrity of authors/institutions could be more readily known… and served as a proxy for judging the accuracy of their words.
But Readers/Receivers are primarily interested in factual accuracy of relevant information — not who wrote/developed it– they need citations that directly validate truth and clearly identify uncertainty in published information. For general Receivers, authors names as the bedrock of a citation-system are unhelpful in a megaworld of information and sources.
Drury, that’s an interesting take but not sure the citation system is past its sell-by date yet! I would say that the primary purpose(s) of the current system are still to justify statements made in one’s paper that relate to prior work and its interpretation, and to provide reference to methodologies employed that the reader might be interested in appraising in more detail.
The citation system also has secondary benefits. It encourages/forces one to read and assess prior work. For example, I’ve got a paper going through publication in a field that’s outside my main research area. We did a load of analyses, got quite excited and wrote a paper before really exploring the prior work in detail. In doing this (especially for preparing our paper Intro), we found that quite a bit of our analysis had been done before (in simplified or alternative form). This was chastening but not a disaster, but it forced us to put our findings properly in the context of the field and to transmit that to anyone that might read our paper – and we learned quite a bit more than we would otherwise have done. And of course all of the authors of the previous work that we knew of from the outset, or discovered during writing, get properly credited by citation. Nothing wrong with that..
It’s difficult to understand in your system how the “factual accuracy of relevant information” is conveyed other than by citing relevant work. Of course some relevant information is either self-evidently correct so no citation is necessary, or is sufficiently paradigmic (if I can invent a word!) that one can refer to a textbook.
And then there’s the question of what we mean by “factually accurate information” and who decides this!
It really depends on the reason for retraction, and the reason for citation. If a paper has really good data but was retracted because the entire introduction was plagiarized, I’d still cite it. If a paper has a really good review or synthesis of a field, but also contains bogus data, I’d also still cite it.
Does this matter very much? Research doesn’t progress by a sort of linear piling-up of findings so that a flawed study scuppers or degrades subsequent work that might cite it! At least in the physical/biomedical sciences a research study is done in the context of a usually large and diverse body of previous work.
It’s also quite possible that a study known (or not) to have been retracted might have a significant influence on the thinking of subsequent scientists – e.g. they are working on the same or related problem and some elements of the ideas or interpretations of the authors of the flawed work might be novel and relevant (and thus citable) even if their paper is sufficiently flawed that it’s retracted.
More problematic is work that is seemingly important but ultimately wrong, and for which its wrongness is not established for some time during which it may cause confusion and fruitless experimentation. There are lots of examples (e.g. the flawed tropospheric temperature measures from satellite microwave analysis that was incorrectly interpreted over a period of 15 or more years, or the flawed stem cell research that falsely purported to describe methodologies for making pluripotent stem cells).
Of course if someone published a paper in which the conclusions/interpretations relied very strongly on retracted research then that would be an embarrassing problem for the author(s).. but probably not for the research field
Mail (will not be published)