Correction, Erratum, Expression of Concern, Retraction: How the Record Gets Fixed

Journals fix the published record with a graded set of notices. A correction (also called an erratum or corrigendum) repairs a specific error while the findings still stand, an expression of concern flags unresolved doubt while a question is being investigated, and a retraction withdraws a paper whose conclusions can no longer be trusted. The label signals severity and certainty, not intent, and a good notice states plainly what changed and who requested it.

Why the published record needs a repair mechanism

Publication is not the end of a study's life. It is the moment the work enters a shared record that other people build on, cite, and sometimes act on. Because no review process is perfect, some of what gets published turns out to contain mistakes, from a mislabeled figure to a flaw that undermines the whole conclusion.

The scholarly system handles this with a graded set of notices rather than a single blunt tool. The gentlest repairs a small error and leaves the paper standing. The most serious withdraws the paper altogether. Knowing which notice is which lets you read the record accurately instead of guessing.

Corrections, errata, and corrigenda: fixing what is fixable

A correction repairs a specific, identifiable error while the paper's findings remain valid. Journals use several names for this, and the labels overlap. An erratum traditionally marks a mistake introduced during production, while a corrigendum marks a mistake introduced by the authors, though many journals now simply call both a correction.

The defining feature is scope. A correction fixes a defined element, a wrong value in a table, a mislabeled axis, a missing disclosure, without disturbing the conclusions. A scholarly analysis of these formats notes that newer variants, such as amendments and retract-and-replace, have emerged to handle problems too large for a simple erratum but not severe enough to discard the work.

Expression of concern: a flag, not a verdict

An expression of concern is a flag, not a judgment. An editor issues one when there is credible reason to doubt a paper but the question is not yet resolved, often while an institution or the journal is investigating. It tells readers to treat the work with caution while the facts are established.

The key thing to understand is that it is provisional. An expression of concern can be lifted if the work is cleared, converted into a correction if a fixable error is found, or escalated to a retraction if the doubts are confirmed. Seeing one attached to a paper is a signal to wait for the resolution, not to assume the worst.

Retraction: withdrawing the conclusions

A retraction withdraws a paper because its central conclusions can no longer be relied upon. Crucially, retraction is about the trustworthiness of the results, not the motives of the authors. Publication ethics guidance is explicit that the purpose is to correct the literature and protect its integrity, not to punish anyone, and that honest error and misconduct can both lead to the same outcome.

A retracted paper is not erased. It stays visible so the record remains complete, but it is marked clearly as retracted, and a notice explains what happened. When only some authors were responsible, a well written notice says so, while recognizing that authorship carries shared responsibility for the work.

What a trustworthy notice looks like

Guidance from the Committee on Publication Ethics sets out what a good notice does. It identifies the corrected or retracted article clearly, states who is issuing the notice, and gives the reason in specific terms rather than vague language. It is linked to the original article in both directions, and it is labeled so that databases and search engines carry the flag wherever the paper appears.

Specificity matters because readers use the reason to judge how much of a field is affected. A retraction for an honest calculation error has different implications from one for fabricated data, and a notice that only says the paper is being retracted, without saying why, fails the people who relied on it.

How to read these labels as a reader

Treat the label as a measure of severity and of certainty. A correction says the finding survived a repair. An expression of concern says the question is still open. A retraction says the finding did not survive. None of the three tells you, on its own, whether the cause was a mistake or misconduct, so read the notice itself rather than inferring intent from the label.

The practical habit is simple. Before leaning on a paper, check whether any notice is attached to it, and if one is, read what it says. The system only protects you when the people downstream actually look.