global English Live English Love English

Colliding with common sense and usage

© GettySome prescriptive usage rules seem so arbitrary and unnecessary as to be made out of whole cloth. One such rule has to do with the word collide, meaning clash or crash into each other, and with related forms like colliding and collision.

According to the rule, you can use these words only when both items in a collision are moving. So if you cycle into a stationary gate, that’s not a collision, but if the gate is swinging at the time, it is a collision. Maybe you find this logical somehow – or maybe, like me, you think it’s awkward and silly. Or it would be, if it were an actual rule.

Collide is also used figuratively (see Macmillan Dictionary’s sense 2 for definitions and examples). This is more common than sense 1, but when you’re talking about, say, opinions or personalities colliding, it has nothing to do with elements moving or not. So the rule is only ever invoked when the word is used in the physical sense.

How did this odd belief arise? Merriam-Webster’s Concise Dictionary of English Usage says it’s a tradition among newspaper editors, and finds it in Hyde’s 1920s Handbook for Newspaper Workers: “Only moving objects collide, and with each other. A car does not collide with a fence.” Collided also appears in the 1877 Index Expurgatorius’s banned list, but without explanation.

Later, Bill Bryson and Theodore Bernstein, among other commentators, adopted the rule. In his Dictionary for Writers and Editors, Bryson insists that collisions “can occur only when two or more moving objects come together. If a car runs into a stationary object, it is not a collision.” He says the same in his Dictionary of Troublesome Words. But in this case it’s the bogus usage rule that is troublesome: it imposes a confusing and unnecessary constraint on writers.

Experts who have researched the restriction dismiss it promptly. Robert Burchfield, in his revised edition of Fowler’s Modern English Usage, says there is “no basis for such a belief” (Fowler’s original work made no mention of it). The Columbia Guide to Standard American English says the proscription is “not true” and that the unrestricted use is standard.

There is no etymological justification for it either. Collide comes from Latin collidere “strike together”, from the familiar prefix com– “together” plus laedere “to strike, or injure by striking”. But all it takes is for one authority to invent a rule, whereupon eager rule-collectors will accept it, repeat it, and proselytise for it, and misinformation will collide with the fear of being wrong.

Email this Post Email this Post

About the author


Stan Carey

Stan Carey is a freelance editor, proofreader and writer from the west of Ireland. Trained as a scientist and TEFL teacher, he writes about language, words, books and more on Sentence first, Macmillan Dictionary Blog and elsewhere. He tweets at @StanCarey.


  • “Collide” seems to be used about accidents between pedestrians and vehicles, even 10-ton trucks / lorries. This seems a bit disproportionate for the poor pedestrian but no doubt it is to do with not apportioning blame or influencing enquiries?

  • Edward: Is it used that way? I agree; that seems somehow unbalanced. If a vehicle hits a pedestrian, I prefer hit. Collide seems better suited to people or non-human objects (trains, celestial bodies, etc.) bumping into each other.

  • Stan:
    When I was a newspaper reporter, we were enjoined from writing that a vehicle hit another vehicle; they always “collided”; that would keep us out of court.

  • Thanks, Marc; that’s a useful point. I’ve heard of similar style guidelines from other journalists. Henry Fuhrmann at the LA Times, for example, told me they used “X and Y collided” rather than “X collided with Y”, to avoid apportioning blame.

  • Don’t get me started on Bryson’s Dictionary for Writers and Editors. I bought it when the OUP work of the same name was out of print, thinking that Bryson and Penguin had somehow taken it over, and swiftly decided it was a complete waste of money. (Apart from anything else, it had about a quarter of the words in it that the Oxford DWE (now NODWE of course) had/has.

  • The car collided with the wall. Sounds good.
    The two cars collided with each other. Sounds good.
    The car and the wall collided with each other. Doesn’t work for me. I guess the subject of the verb must be moving.

  • Harry: I remember reading The Mother Tongue, Bryson’s general-readership book on language, many years ago, and though it’s enjoyable in many ways, it also has a lot of mistakes.

  • Peter: “The car and the wall collided with each other” could work, but only if the wall was also moving. (It’s probably happened in an action film.) If the subject wasn’t moving at all, a collision would be unlikely.

  • To be consistent with the sprirt of this “rule”, surely it is not enough that both protagonists be moving, but also requires that some component of their velocity must be in opposition? To put it another way, the silliness of this rule is emphasised by considering that if a car hits a bicycle that is moving in the same direction, it is a collision, but not if the bike is stationary.

  • Niall: That’s another thing I’m not sure about, since I never obeyed the pseudo-rule or heard it explained in detail. To take another example of your “opposing velocities” point, as I understand it: If a person running in a tight race bumps into the person in front and they fall, is that a collision?

  • Stan: for an example of a tight race and a “bumping into the person in front”, see the women’s 3,000 metres at the 1984 Los Angeles Olympics: The relevant incident starts at about 2:04. I’ve looked at several accounts, and many talk about the athletes “bumping into each other”, but there are plenty of reports which use the words collision or collide. The Guardian in 2012, for example, captioned a photo of the incident: “America’s Mary Decker plunges to the infield grass after a collision with Zola Budd”. Using the verb, RunnerDude’s blog seems to blame Decker: “…fellow runner Mary Decker-Slaney collided with Zola during the 3000m race” while The Running Review makes Budd the subject of the action: “Budd … collided with crowd favourite Mary Decker of the USA”. To avoid pinning the blame, says “Budd and American track star Mary Decker collided in the 3000m”.
    The OED hasn’t revised its entry for collide since 1891 (what do they do all day there?) and gives no guidance about the need for both parties to be moving. For a definitive ruling on this one, we might have to turn to the Greatest Living Grammarian, Nevile Gwynne 🙂

  • Stephen: Thanks for this example. It’s really interesting how various reports structure it differently: X collided with Y; Y collided with X; X and Y collided. Suggestions of blame aside, it’s clear at least that collisions can occur when both bodies are moving in exactly the same direction.

  • Some dictionaries call this type of verb reciprocal verbs, because of the patterns you describe, Stan. Other examples are meet, argue, kiss, hug, touch, talk, and some others I won’t mention here. The implication is that the parties involved are doing the same thing to each other or participating in the same event. So if Jane argues with Dave, it’s not just Jane sounding off, Dave has things to say as well. If Jane and Dave then hug, Jane hugs Dave and Dave hugs Jane; both parties are both hugger and huggee.

  • Liz: I remember encountering a peeve about the everyday use of the phrase talk to, as in “I’ll talk to you later” or “I was talking to her yesterday”. The stickler said it should be talk with, because both parties were talking; it wasn’t a monologue. But the phrase implies reciprocity, and that’s how people ordinarily interpret it. Reciprocal verbs is a handy term for this, thank you.

  • There’s a whole chapter on reciprocal verbs in the, frankly rather fabulous and under-circulated, Collins Cobuild Grammar Patterns series (1: Verbs, Harper Collins 1996). Lots of corpus-based data and a full and in-depth analysis of syntactic patterning – just one of Stephen et al’s many legacies… . 🙂

  • and how weird – now my comment disappeared when I made a comment about the typo in my name…

    So, what I meant to say was that there’s a whole chapter on reciprocal verbs in the (frankly rather fabulous but very under-circulated) Collins Cobuild Grammar Patterns series (1: Verbs, HarperCollins 1996), The chapter includes a raft of corpus attested data on these verbs and an in-depth examination of syntactic patterns, just one of Stephen et al’s many legacies … 🙂

  • Thanks for the hat-tip, Kerry, but the Grammar Patterns books were entirely down to Gill Francis and her co-authors Susan Hunston and Elizabeth Manning. If teams of grammarians were car designers, they’d be Charles Rolls, Henry Royce, and Enzo Ferrari.

  • Thanks for the book tip, Kerry. The Grammar Patterns series sounds great, and I’ll keep an eye out for it. (Your comment about the typo appeared for me before the earlier one. I figure the spam filter held “Kerr’s” comment for moderation because the name didn’t match a previously allowed ID.)

  • Stephen and Kerry – Thanks for your nice comments on our little-known Grammar Patterns series. If anyone’s interested, the whole of the first volume, ‘Verbs’ can now be browsed online for free – which is the best place for it as it is a reference book – at

  • A very late comment – I just found this blog after being steered here from elsewhere (and avoiding a collision!).

    In nautical terms, a ship may collide with another ship, but a ship hitting a stationary object such as a dock is said to allide with it. The distinction is just what this post discusses: stationary vs moving. Seems a useful word, but apparently lost as so many sailing terms have been.

  • Late comments are just as welcome as old ones, Jim! Allide is an interesting word, and unusual too: I don’t think I had come across it before today. Many sailing terms have indeed faded from use, while others have been adopted in figurative use, as I noted in a recent post.

  • While lost to general use (if it was ever found there), it’s still common in maritime circles, as you can see by searching Google for “ferry allision”. You’ll find several US Coast Guard reports, etc. especially relating to a Staten Island Ferry accident of about 10 years ago. That was when I first came across it, in a New York Times article on the accident. I suppose the writer just picked it up from the USCG report. but at the time I was impressed, and quite pleased to encounter a new word. (-:

Leave a Comment