Intended for healthcare professionals



BMJ 2000; 320 doi: (Published 08 April 2000) Cite this as: BMJ 2000;320:1005

This article has a correction. Please see:

What happened to the false allegation in

  1. Brian Morgan, freelance journalist (brianmorgan{at}
  1. 4 Rawden Place, Riverside, Cardiff CF11 6LF

    EDITOR—Several months ago I received an email from a reader puzzled by a reference that I had made in an earlier rapid response1 to Marcovitch's article on doctors who have exposed child abuse2 [a shortened version of that earlier response is published here as the third letter in the previous cluster, p 1004]. I had written that Marcovitch had made a false allegation of perjury, but the reader was puzzled because there was no such false allegation when he came to read the response.

    Soon after I wrote my response a “rapid correction” was made and a retraction of the allegation posted. The reader was not to know this, for soon after that, and before he read the response, the allegation was deleted from the text in and the correction was also deleted. The correction appeared in on the same day that Marcovitch's article was published in the paper journal. Any readers who see only would be puzzled because both Marcovitch's allegation and's correction have been deleted from

    The allegation and the correction appear in the printed BMJ, which must be assumed to be the definitive text. The trend now is towards electronic versions of biomedical journals having different content from the printed ones, by design, and there being online journals with no printed version. Is there no one who shares my concern that internet publications should have the same permanent authenticity as printed ones?

    If publishers may remove text, or even add text, what evidentiary value do such documents have as source materials?


    1. 1.
    2. 2.

    Editor's reply

    1. Richard Smith
    1. 4 Rawden Place, Riverside, Cardiff CF11 6LF

      Morgan raises an important question. It is clearly impossible to correct errors in the paper journal after publication by correcting every copy. It is, however, possible to correct by deleting the error or correcting it. But should we do that?

      We have decided not to in general. Instead, we will correct errors as we usually do by publishing a correction. That correction is then linked to the piece it is correcting, so that anybody reading it will be aware that there is a correction. The electronic medium does thus have an advantage over the paper medium.

      We have adopted this policy to leave a trail of errors and corrections. They may well prove to be important at some future time. There is something unnerving—and totalitarian—about “rewriting” history.

      An exception to our policy is when a lawyer tells us to remove something, which is what happened in the case that Morgan is writing about. Perhaps lawyers and the law will eventually take a different view.

      View Abstract