You almost had to feel sorry for Gordon Brown this week, frozen into silence by the question whether he approved of the decision to release the man convicted of the Lockerbie bombing.
He compounded the impression of hapless impotence by committing a common error of usage. Mr Brown declared himself "both angry and repulsed" at the jubilant reception the freed man received in Libya. The Prime Minister was not repulsed but repelled.
The two words are derived from different parts of the same Latin verb, repellere, meaning to push back. The English words have diverged in meaning. "Repulse" is found in military contexts, meaning to drive back by force. If you mean affect with disgust or aversion, the right word is "repel".
Power of suggestion: On Tuesday we reported on a scientific study which found evidence that women with high levels of testosterone may tend to pursue high-risk careers in business. The report was accompanied by some photographs of high-achieving business women. One of the captions read: "Carol Bartz, 60, CEO of Yahoo. Known as a 'tough operator', in 1992 she survived a bout of breast cancer."
No sane person would ever write: "This woman survived cancer because of her tough personality." State it outright and anybody can see how ludicrous the idea is, and how likely to annoy cancer patients. But this caption keeps a straight face while implying it.
Hands off: People often refer vaguely to any document of great age as a manuscript. M M Deyes writes in from London SW5 to draw attention to this, from a news story published last Saturday: "The missing volume, printed in 1623, was among a number of manuscripts taken from Durham in the raid."
If it was printed, it isn't a manuscript. The Latin derivation of the word could not be clearer: manu – by hand; scriptus – written. A manuscript is a document written by hand.
Cliché of the week: If it's got palm trees and sandy beaches bordering the Caribbean Sea, it must be an island paradise, right? That seems to have been the thinking of those responsible for the headings and picture caption that accompanied a report yesterday about Lord Ashcroft, the Tory donor, who lives in Belize.
"Paradise in Belize turns sour for Ashcroft," declared the main headline, though the story offered no evidence that the country is anything like Paradise. And a picture caption identified the Prime Minister of Belize "who has accused Michael Ashcroft of subjecting the island to 'new-age slavery'". That appeared directly above a map showing that the former British colony is, in fact, a small piece of Central America, firmly anchored to Guatemala and Mexico.
True enough? Headlines are supposed to be as interesting as possible, but still true. You are not allowed to make up the best headline you can think of that would fit the facts if the facts fitted the headline.
Last Saturday we ran the story that Ken Livingstone, the former Mayor of London, is to marry in a ceremony at the Mappin Pavilion, a function space at London Zoo. Readers were reminded that Livingstone is an animal-lover, once known for keeping newts. They were told that the Mappin Pavilion is next to the reptile house.
The headline was "'King Newt' Ken to marry at the reptile house". Two problems: newts are not reptiles, so the relevance is obscure; and the words "marry at the reptile house" mean that the ceremony will take place in the reptile house itself, not in a building next door. So the headline, unfortunately, is just not true. All-but-true doesn't cut it, I'm afraid.
Rough estimate: Whenever you write "overestimate" or "underestimate", pause for a moment to make sure you have picked the right one. Thursday's Big Question was about Wikipedia. "And given that it is one of the most popular websites in the world its utility is difficult to underestimate."
The bigger a thing is the easier it is to underestimate, because more of the available likely-sounding estimates of its size will be underestimates. Compare, say, a man five feet tall with a man six feet tall. All figures between five and six feet will be overestimates of the shorter man's height and underestimates of the taller man's. So it is easier to underestimate the height of the taller man.