We use metaphor, a figure of speech, to explain or enliven: in doing so we write metaphorically, or figuratively. The opposite of metaphorically is literally. We don’t need to add metaphorically to a metaphorical statement; we rely on readers to recognise the metaphor: “Doctors’ morale has hit rock bottom,” not, “Metaphorically, doctors’ morale has hit rock bottom.” But oddly, we sometimes feel the need to add literally to bolster metaphorical statements: “Doctors’ morale has literally hit rock bottom.” This use of literally has long been discouraged in style guides—for good reason. Long before Michael Gove MP issued his advice to civil servants, the Treasury invited Sir Ernest Gowers to help them to improve official English. The result was The Complete Plain Words (which combined two of Gowers’ earlier works), published in 1954 and recently revised and updated by his great-granddaughter. Parts of Gowers’ original works are on the internet, where you can read his examples of this “foolish use” of literally: “Miss X literally wiped the floor with her opponent,” and, “M. Clemenceau literally exploded during the argument.” Bill Bryson advises that literally be used only when something expected to be metaphorical is actually true: his example is, “He literally died laughing.” If he really did.

There are a fair number of literallys in PubMed, and most of them are not needed. It is enough to write that recognising type I diabetes in children can be life-saving, that there are hundreds of fungicidal polypeptides, and that the purpose of a shortening osteotomy is to shorten the bone: literally has no place in any of those statements.

Literally is appropriate in describing autophagy (which allows the recycling of cellular components) as “literally self-eating,” and almost correct in describing interviews as being “transcribed literally:” word-for-word or verbatim is better. It is certainly correct—in the Bryson sense—in an obituary of a man who “penned—literally, since he never used a keyboard—and edited more than 30 books…” I can’t think what literally means in “randomly selected, literally healthy subjects, of similar age…,” but suspect it is the common problem of words subtly changing their meanings because of incorrect use by writers for whom English is an additional language.

Literally is most appropriate in the considerable body of research into the comprehension of metaphors. Anyone can have problems interpreting metaphors in another language, but patients suffering from some psychiatric or neurological illnesses have difficulties in their own language. There are studies of how the normal brain responds to different figures of speech and to the implied meaning of the speaker. To give a flavour of cognitive pragmatics, as the field of study is known, participants in one study, using functional MRI, were presented with four pairs of target sentences that were metaphor versus literally coherent, metaphor versus literally incoherent, sarcasm versus literally coherent, and sarcasm versus literally incoherent.

We have known for decades that there are cells in the occipital cortex that detect edges moving in a particular direction. Perhaps there are also cells that detect inappropriate metaphor used sarcastically in a literally incoherent sentence.

Competing interests: I have read and understood BMJ policy on declaration of interests and declare that my only competing interest is my co-authorship of a book about medical English.

Neville Goodman is a retired consultant anaesthetist and a writer, and co-author of a book on medical English.