David Bradley opens his Nature Matters post with a recollection from his student days, when a friend on an engineering course would pad essays with deliberately overwrought prose to meet word counts, a practice he likens to today’s reliance on language models to generate filler text. He then explains that the phrase 'vegetative electron microscopy' is not a real technique but an artifact of an optical character recognition (OCR) error in a 1959 two‑column journal article. In the original paper, the left‑hand column ended with 'vegetative cell' while the right‑hand column began with 'electron microscopy'; a line break caused the OCR system to merge the end of the left column with the start of the right column, creating the nonsensical hybrid phrase.
Because a human reader would immediately see the mistake, the appearance of 'vegetative electron microscopy' in a modern manuscript indicates that the text was copied without proper checking. Bradley cites Retraction Watch, which found the phrase in numerous papers over recent years, noting that it has become a telltale sign of plagiarism: researchers who scrape boilerplate content containing the OCR artifact and paste it into their own work fail to notice the gibberish, thereby revealing a lack of substantive scrutiny. He adds that the term’s presence shows that neither the authors nor the editorial team nor referees read the manuscript before submission.
The scientific community has responded by treating the phrase as a plagiarism fingerprint. Bradley reports that some articles containing it have been retracted, others have received corrections, and a few remain in the literature despite publishers’ apparent indifference—he notes that some publishers seem to care only about collecting fees and not about editing out such drivel. This persistence underscores broader weaknesses in academic publishing, where inadequate checks allow nonsense to enter the scientific record, potentially eroding public trust in science and giving pseudoscience purveyors leverage to claim that science itself is untrustworthy.