When did “by” become “after”?

This post is by Phil Price, not Andrew.

I just did a Google News search for “injured after”, and these are some of the headlines that came up:

16-year-old bicyclist seriously injured after being hit by car in Norfolk
At least 1 injured after high-speed crash in Bridgeport
Teen injured after falling off rooftop
Driver injured after crash down embankment near Sherwood
Two injured after shooting on Indy’s east side

There are many more like these. They all irritate me. What, the 16-year-old cyclist was uninjured in the crash, but after the crash he got hurt somehow? The high-speed crash in Bridgeport didn’t injure someone, but someone got injured afterwards? The only one of these that I believe could be factually correct is “Teen injured after falling off rooftop”, since, yeah, ha ha, it wasn’t falling off the rooftop that hurt him, it was hitting the ground a couple of seconds later.

These should be “16-year-old bicyclist seriously injured by being hit by car”, “At least one injured in high-speed crash”, “Teen injured by fall off rooftop”, “Driver injured in crash down embankment”, and “Two injured in shooting”.

I first noticed this factually incorrect use of ‘after’ a couple of years ago but it was pretty uncommon. Now it seems to have taken over, or at least it seems to have caught up with “by” and “when”, as in ‘injured by crash’ and ‘injured in crash’ for example. And don’t bother telling me Shakespeare used ‘after’ this way, or Austen or Chaucer or Milton, it’s still wrong.

My friends, including my writer friends and editor friends, tell me to get over it, what’s the big deal, you know what they mean. Well, first of all I don’t always know what they mean, there have been times I’ve been genuinely unclear on what is being described. But also, just because I understand it doesn’t mean it’s right. If I say “this is just not the write way to phrase it”, hey, you know what I mean, but that sentence is still wrong.

I may be the only one who cares, but by god I am not giving up this battle. This new usage stinks.

Now all you kids get off my lawn.

This post is by Phil.

97 thoughts on “When did “by” become “after”?

    • Thanks, Shravan! Somebody had to make that joke.

      The right answer is what Ignato Fiorentino said. Let me unpack that a bit. What’s going on is the grammatical distinction between using “injured” as a transitive verb (A injured B) and as a stative predicate (B is injured). The second usage is not the passive form of the first (B is injured by A). The passive form is a present-tense statement about the act; the stative form is a present-tense statement about the condition. A clear stative example is “Our veteran starting pitcher is injured, so the rookie is getting his big chance.” That sentence is not narrating the act of the injury in the present tense, it’s just describing the state of the starting pitcher in the present tense. Language is fundamentally overloaded.

      What is annoying about the examples to me isn’t that the usage of “injured” is incorrect, it’s that it disavows the author of making the causal connection. So maybe it’s a bunch of nervous lawyers calling the grammatical shots.

      Greg Tucker-Kellogg gives a great example below, too, contrasting with the stative “uninjured” and Thanatos Savehn identified the causality issue. I don’t think the issue is that prepositions are idiomatic or that they’re evolving (two different though related concepts). They’re highly variable semantically across languages because they’re kind of random groupings of often spatial or temporal properties, but this is about the stative usage, not the preposition, which is being used in its traditional ordered form. Joe Nadeau suggest reordering to “car seriously injures 16-year-old bicyclist”, but that changes the focus of the sentence to the car from the 16-year old and thus changes the narrative smoothness. Ignato also mentions that in his summary. Also, as Betty Barcode writes, do you want to pin it on the car or on the driver if we’re going after causality?

      Among Phil’s examples, I do like “injured after falling off rooftop”. That’s just good semantics, because falls don’t usually injure people, breaking falls does.

      P.S. +1 for more posts that overlap with my early career work on the syntax-semantics connection!

      • I have to confess to playing a little trick by using the ‘injured’ example. Search Google News for ‘killed after’ and you find the same construction. The state of being dead lasts forever, but ‘killed’ refers to an act or moment that comes to an end. ‘Teen killed after being hit by truck’ only makes sense if someone strangled her on the way to the hospital.

        • Teen has been killed after being hit by a truck… This is really just the same thing as the fall example. Technically the person probably died at some moment after the initial contact by the truck… so if you like the fall example but you don’t like this one, it comes down to us narrowing in on your “perceptual atomic time tick” what “is at the same time” for you.

          I think it’s just sensationalism here… as “Teen dead after being hit by truck” is less dramatic sounding.

        • Like I said, Teen Killed After Being Hit By Truck is fine with me as long as it wasn’t being hit by a truck that killed her. If being hit by a truck was what killed her, even if not instantly, then she was killed by being hit by the truck, she was not killed after being hit by the truck.

          ‘Teen Dead After Being Hit By Truck’…well, I don’t like it because it lacks causation as Bob points out, but at least it’s not wrong.

          Teen Killed After Being Hit By Truck, no, I don’t see how that can be right…unless, as I said, someone strangled her on the way to the hospital.

        • Phil’s text, my copy-editing for mention quotes:

          “Teen Killed After Being Hit By Truck” is fine with me as long as it wasn’t being hit by a truck that killed her.

          I don’t think the verb “killed” is typically used as a predicative adjective. So while “He is injured” sounds fine, “He is killed” sounds strange. You often find that when there are good, high-frequency words for a task that you don’t find regular alternatives. In this case, “dead” is a common word used for the state one is in after being killed, whereas “injured” itself is the word we use for the state one is in after being injured.

        • Yes, exactly right. ‘Dead after’ is be correct, we are referring to the state of being dead rather than the act of becoming dead, but ‘killed after’ definitely refers to the act rather than the state. ‘Teen killed after car accident’ is a bad headline. But this construction is common now.

        • Well, initially it wasn’t a trick, I started with the ‘injured’ example. I’ve seen this ‘after’ construction in many contexts, not all involving accidents or anything in particular, but the one that prompted this post was a headline, ‘Six pedestrians injured after tree falls’, which made me think What, did they all trip over it?

          But as I was writing, I realized ‘injured’ could refer to the state of being injured, not the act of being injured, so it’s not technically incorrect. Rather than rewrite the whole thing — I had already pasted in the headlines, which are real examples — I decided hey, this could actually be sort of fun, because a bunch of people are going to say ‘No no, Phil, it’s not incorrect because those people are still injured, so they are injured after the tree fell!’ and then I get to say “What about ‘killed after being shot’, smart guy?”

          So it was either a little trick or I was too lazy to rewrite. You can take your pick.

        • then I get to say “What about ‘killed after being shot’, smart guy?”

          You can say it, but the problem is that “killed” doesn’t function well as a predicative adjective. Did you find instances of “killed” being used in the same way as your “injured” examples?

          I cover the whole concept of these flexible lexical relations in my book, Type Logical Semantics, in section 11.4 and in the related exercises. It was supposed to be chapter 6 of my Ph.D. thesis, but my advisor told me I was done after 5 chapters, and it took another 10 years for the non-computational issues to find a home.

        • Yikes. I should’ve looked first. There are a lot of hits for the query [killed after *] on Google. That I do find strange given that it’s hard to understand “killed” as being a predicate. If one wanted to nitpick tense, one might argue that “Pilot killed after plane crashes into home, bursts into flames in New Jersey neighborhood” that the killing happened after the crash and flames the same way the injury happens from the landing, not the fall. But I don’t think that’s what’s going on here. It is what’s going on with “Hunter shot, killed after friend mistook him for deer in Georgia”.

        • Bob, I told you ‘killed after’ is in common use! That was the ‘little trick’ I played by being lazy (see response to Carlos): I realized ‘injured after’ can be correct but ‘killed after’ cannot, but left ‘injured’ as my example because I was too lazy to fix it, realizing some people would fall into the trap of defending it.

          BTW ‘after’ used in this way doesn’t just come up in the context of injury or death, it has somehow become acceptable or perhaps even preferred where ‘by’ or ‘when’ would be either more correct or more appropriate. Now that I’m sensitized to it, it bugs me even in innocuous cases. Team Wins After Scoring Five Touchdowns. Sure, ok,one can argue that no matter how far ahead they were, they didn’t win until the game ended and that was necessarily after all of those touchdowns. But why not Team Wins By Scoring Five Touchdowns, or Team Scores Five Touchdowns, Wins. Or Team Scores Five Touchdowns in Victory. (As I said, I’m aware ‘after’ is fine here, I’m so irritated by it in general that I have stared to hate seeing it).

      • Thanks, Bob. More seriously though, I think Phil is completely wrong. There is an implicature that the injuring event happened long after or immediately as a consequence of the other event, but there is no logical entailment. One can check this:

        Consider:

        16-year-old bicyclist seriously injured after being hit by car in Norfolk

        One can confirm the implicature that the injury happened long after:

        16-year-old bicyclist seriously injured after being hit by car in Norfolk: she escaped relatively unscathed from the car accident but her bike was totaled. Later on, she was hit by a bus while walking home, having abandoned her bike.

        Or one can cancel the implicature:

        16-year-old bicyclist seriously injured after being hit by car in Norfolk: the car hit her head-on and she broke her arms and legs.

        Phil is making the mistake of taking only one of the interpretations as the only one that’s possible. To take an extreme example, it would be like interpreting the sentence

        The girl saw the boy with the tennis racket

        to only have the interpretation that the seeing event was carried out by using the tennis racket as a telescope (obviously an absurd thing to do, but syntactically fine), and not that the boy was holding the tennis racket. Compare with the truly ambiguous case:

        The girl saw the boy with the telescope.

        Did the boy have the telescope or the girl? Both readings are admissible. It’s the same with the after examples.

        That’s my take on it anyway.

        • It seems to me that you’re agreeing with me, not disagreeing! Teen Killed After Car Accident makes it sound like the teen survived the car accident and was shot on the way home, or something. If the teen was killed by the car accident that is a bad headline.

        • No, I think he’s saying that the length of time is a free variable that you are to infer… and that you are insisting on inferring that it should be more than at least tens of minutes, whereas others seem to be ok with maybe tens of microseconds…

          “boy killed after car accident”… car accident initiated at time t=0 and by time t=0.00001 seconds the boy was killed.

          vs

          “boy killed after car accident”… car accident initiated at time t=0 and finished at time t=4s, boy has some missing time, and at time t=3320s the boy was suddenly killed by strangulation or something…

          Most people are ok with not inferring 2, in the same way that “seeing the boy with the tennis racket” people are ok with NOT inferring that the girl somehow converted the tennis racket into an observation instrument that allowed her to see better…

        • “time flies like an arrow” for most people doesn’t mean that they should get out their ballistic chronographs and try to calculate the velocity of fruitflies in the same way they’d calculate the velocity of an archer’s missiles… or that special flies from the world of Dr Who (time flies) enjoy buzzing around those archer’s missiles…

          language always has ambiguity and people choose what they think the most logical meaning is… it’s not always clear, but “time flies like an arrow” is virtually unambiguous to the average person, as is “boy killed after car accident”

        • Yes, exactly, that’s what I was getting at. There are ambiguities everywhere in language, and there is no requirement that we disambiguate when talking, because people accommodate the intended meaning. There are lots of examples of this.

          If I say: It was John who bought a book” I strongly imply that only John bought the book. But I can cancel that implicature by continuing: It was John who bought a book, in fact, Mary did too.

          By contrast, if I say: Only John bought a book, now there is a logical entailment. I can’t go on with: Only John bought a book, in fact, Mary did too. (Unless I am actually taking back the assertion I made starting with Only John…).

          When we use the cleft (It was John) the reader may draw that implicature (that it was only John who bought the book), but doesn’t need to (and probably every linguist I know would be spinning his/her wheels: but did someone else buy a copy of the book too?).

          We have a paper on this topic: Heiner Drenhaus, Malte Zimmermann, and Shravan Vasishth. Exhaustiveness effects in clefts are not truth-functional. Journal of Neurolinguistics, 24:320-337, 2011.

          http://www.ling.uni-potsdam.de/~vasishth/pdfs/DrenhausZimmermannVasishthinpress.pdf

          It’s a comparable situation here. Phil is insisting on choosing only one of the two possible meanings. Phil needs to chill :)

        • I thought I put this here, even though it doesm’t exactly fit. But it relates to the idea about ambiguity due to multiple senses of “after.” If you really want to know the answer to when “after” acquired the sense “by”, the answer is as long as English has been spoken. In Beowulf (year is debatable, but say 700-1000 AD) you find:

          Þa þæt sweord ongan æfter heaþoswate hildegicelum..wanian.

          Which means something like, “ the sword began to decay after battle sweat.”

          “after” has a lot of different senses. The meaning that is probably relevant here is “as a consequence of” It’s somewhat related to “would you trust him again after what he did?” , where the temporal meaning is really secondary so there isn’t the same ambiguity

          I’m not sure if this is new or not. It might very well people, but observations about language change are almost always untrustworthy. I think the term is “recently illususion”

        • Yes, yes, the “you know what they mean” argument. As I’ve said a few places, I do know what they mean. But I still see no advantage whatsoever to use “Boy Killed After Car Accident” instead of “Boy Killed in Car Accident” or “Boy Struck and Killed by Car”. “After” implies a temporal separation that isn’t there, as well as minimizing the causality of the incident.

          I understand what you guys are saying, and you understand what I’m saying, so we’re going to have to agree to disagree. But for the record I feel as strongly as ever that this construction is bad.

        • “I feel as strongly as ever that this construction is bad.”

          “bad” is a relative term, and too strong here, IMO. How’s this for bad:

          Boy the met match his.

          Now that’s something I’d feel confident calling “bad”. :)

        • “Boy the met match his”

          Well, the Basque spelling is a little funny, and I can’t see the agreement morphology, but otherwise it seems ok…

  1. Stay strong, Phil.

    My own guess is the usage derives from the (often much more newsworthy) “uninjured after”, e.g. (also from Google News):

    Student Uninjured After Making Contact With Moving Bus
    Filipino human rights lawyer uninjured after gunmen ambush car (this is an especially good headline because it includes a photo of a bullet-riddled car)
    Man uninjured after truck crashes into pond

    In all of those cases “by” would not work but “after” does. After (or perhaps by?) enough of those headlines, “injured after” seems a minor change to headline writers and copy editors.

  2. Interesting. As a non-native speaker, I always thought these sentences were meant to emulate the breathless telegram just-the-most-necessary-words style by omitting the verb in the first part of the sentence, which describes the current state after some event:

    16-year-old bicyclist [is (now)] seriously injured[,] after being hit by car in Norfolk
    Teen [is] injured[,] after falling off rooftop
    Two [are] injured[,] after shooting on Indy’s east side

    What about this?

    Seventy-two [are (now)] dead[,] after bomb explosion on cruise ship

    In this example, replacing “after” with “by” sounds weird to my ears.

    • Yes, ‘injured’ is a state that persists after the injury occurs. So is ‘dead’ for that matter. But ‘killed’ is not, it refers to something that happens at a specific time…and this same ‘after’ construction gets used for killed. Gets used for a lot of things where it doesn’t make sense. It’s just a new use of ‘after’ to imply causation.

  3. “Man better after falling from window.”

    Yes. Conflating “by” and “after” can cause real confusion. If we didn’t conflate the two this headline would be clear … the man is now better due to medical care he got after he fell from the window. But, because we conflate the two, we can’t tell … the fall from the window may have made him better.

    (This example is at least twenty years old, so this has been going on a long time.)

  4. As someone fascinated by word/language usage and ambiguity thanks for this bit I’d never thought about. With that said though, I do believe language not only evolves, but is very largely idiomatic — which is to say that the meaning of phrases and word combinations, through usage, can be rather different from the literal meaning of the individual words. So not something I’d tilt at windmills over.
    (Hey, but don’t get me started on people who use “Woah” in place of “Whoa”! ;)

  5. Instead of “16-year-old bicyclist seriously injured by being hit by car”, how about “car seriously injures 16-year-old bicyclist” – avoiding the passive and ambiguous phrasing entirely.

  6. Hey, on a related note, check out the #DriverNotCar hashtag on Twitter. It’s an effort to get journalists to replace “A pedestrian was hit by a car” with “A driver struck a pedestrian.” It’s important to attribute agency correctly in crash reporting.

    • “A driver struck a pedestrian.”

      I’ve got a problem with this. The driver did not strike the pedestrian, the car did.

      It also seems an open license for reporters to insert editorial comments about “the root causes” of events, such as “culture of left-wing permissiveness kills child [that fell out of window].”

      Also, how do you know the driver didn’t have a heart attack and die?

      • You could say “Driver injures pedestrian” though. If someone uses a weapon to injure someone, it’s generally accepted that you can put that on the perpetrator rather than the weapon. You can say ‘Mugger injures victim’ even if he used a knife or a gun to do it, for example.

        • Did Boeing kill the LionAir victims? Or did the plane do it? Or did the pilots do it? Or was it an accident? Or the MCAS system? The sensors? Dennis Mulenberg? Boeing shareholders? The Machinists? The designers?

        • Here’s another instance of how subtle language is.

          I was objecting because of the word “struck”. Yes, it is correct to say the driver struck the pedestrian (with a car), but it also suggests the driver struck the pedestrian with his fist, so saying “a driver struck a pedestrian” is confusing.

          Even “driver injures pedestrian” has the ambiguity that it isn’t clear the driver did it with a car. He might have shot the pedestrian.

          “Pedestrian struck by car” seems clearest. It strongly implies that the car had a driver. (If not, that fact would have been in the headline: “driverless car strikes pedestrian”.)

        • “It strongly implies that the car had a driver.”

          Today, at some point, highly likely; but in the future maybe not! Angry cars might go striking pedestrians of their own accord.

  7. Wonder if it’s a legal / fear of being sued thing? Like, “I didn’t say you assaulted a man with your car. There was just a car and mysteriously now the person is hurt.”

  8. You may not like this usage, but it’s not wrong. What you are missing is that in these sentences, “injured” is operating as an adjective, not a verb. Compare: I am tired after staying up all night. In the sentences you prefer, it would be operating as a verb [and would be in the dreaded passive voice].

    This is called the past participle. For more information, see the Student’s Introduction to English Grammar. at http://www.lel.ed.ac.uk/~gpullum/SIEG/

    It could be that the usage is increasing because newspapers think their readers are now more interested in the person’s condition after the accident [and less interested in what happened in the moment of the accident] than they used to be, but I’m guessing that the adjectival usage is more frequent because of popular agitation against using the passive voice. So I suggest that you get together with the anti-passive-voice fanatics and have it out.

    • On the one hand, yes. On the other hand, no. A word means whatever people think it means…but if you use it to mean something people don’t think it means, it’s wrong. And if you use it in a way that is ambiguous, it may be ‘right’ but it’s still a poor choice. “Yeah, O.K., but you know what I meant” means you made a poor choice.

      Teen Dead After Car Crash, O.K., fine, I would accept that although I’d hope for less ambiguity about what killed the teen. Teen Killed After Car Crash, no, unless someone strangled her on the way to the hospital she was killed by the car crash or in the car crash or when the car crashed.

      I intend to fight on this hill as long as I can. It’s my new hill after having to finally abandon ‘data are’. But the latter was purely convention. This ‘after’ nonsense is just wrong.

      • I read your example as:

        Teen Killed After [Involvement in] Car Crash

        I guess it might be better

        Teen Dies After [Involvement in] Car Crash

        But it’s less sensational… I mean, people die after getting old, or after struggling with hear failure… but “killed” implies some agent caused the event and therefore some drama.

        • But in this example the teen wasn’t killed after involvement in a car crash. It’s the car crash that killed her. I’d accept ‘Teen Dead After [involvement in] Car Crash’, because that is simply true. But ‘killed after’…I don’t think that works and it doesn’t matter how long the interval between the crash and the death. If someone is shot in the head and lingers for five days in the hospital before succumbing, it would be correct to say they died after being shot in the head, but not that they were killed after being shot in the head. Bowing to the people who say there are no real rules in language, I guess the most I can say is that the latter construction seems wrong for me for more than one reason, and does not seem right to me for any reason.

        • It’s an interesting issue that comes up in Detective Fiction (and obviously probably real life) occasionally. And since Detective Fiction is on the list of accessory topics on this blog I’ll continue ;-)

          So, in one of the Harry Bosch novels there’s a character who was shot in some sort of apparently gang related event at a park several years earlier, and then something like 10 years later dies of complications from the wound (the bullet was embedded near the spine if I remember, it’s been a number of years since I read the novel and don’t remember which one it was so I can’t google the details).

          So literally, the victim was killed after being shot by a sniper. It’s not just that he died, but that the cause of death was attributed to a particular person’s violent assault. Until he died, it was just Assault with a Deadly Weapon, or Aggravated Assault and Battery or something. But as soon as he died of complications from the assault… it became a murder investigation.

          If you were in the process of filming some movie stunt on high-speed camera, and something went horribly wrong, you could say “at frame 228595 the victim was struck by a falling beam from the collapsing scaffolding, and at frame 229101 the victim dies when their head disconnects from their body…” gruesome, but the point is that Doc Edgerton would probably argue one always dies *after* some event… and you aren’t killed until you die… it’s all a matter of time-scale.

        • I see your point but you are really splitting hairs.

          To split some hairs myself:
          First, although it’s true that one always dies after _some_ event, it is not true that one always dies after the event that is being reported. If someone is in a violent car crash and dies before the wreckage has even stopped moving — decapitated, for instance, to use your example — it would not be true that they died after the car crash. But at least some reporters would still report it that way. It seems to me that in at least some of these uses the word ‘after’ is not a temporal reference at all, it really does just replace ‘by’ or ‘due to’ as an indication of causation.

          Second, although it might seem inarguable that one is not killed until one dies, you underestimate me if you think I can’t argue the contrary. If you shoot someone in the head and they die on the way to the hospital, I think we would agree that you killed them. But when did you kill them? Did you kill them when you shot them, or did you kill them when they died? I think we would agree that you killed them by shooting them. But if you killed them by shooting them, didn’t you kill them when you shot them? I would say you killed them when you shot them, but it took them a while to die.

          In any case, no matter how many hairs you or I split, you’re not going to get me to accept “Teen killed after car accident” if the car accident killed the teen.

        • Causation really is a tricky thing. When person A shoots person B, it’s the case that A pulling the trigger caused the death, but so did A aiming the gun accurately, and so did A deciding to pull the trigger, and so did the bullet entering person B’s body, and so did maybe the infection they got in the hospital…

          If you look for a root cause, you can argue maybe that the root cause is that person B slept with person A’s spouse… or maybe it’s that person A learned about person B’s affair from person C, or maybe it’s that person A’s adrenal gland secreted excess adrenaline after learning of the affair.. or maybe it’s that as a young child person A had to deal with the fallout of their own parents having an affair, or maybe it’s that person A’s zygote was fertilized by an XY sperm instead of an XX…

        • From the journalists perspective perhaps they really are seeking to avoid causation claims, one thing just happened to occur after another thing… we can report that “objectively”… Although the death didn’t necessarily happen after the completion of the car crash, it did happen after the initiation of the car crash… so it’s still technically correct:

          Bicyclist killed after [initial contact during] car crash…

  9. Phil,

    I’d recommend the book “The Unfolding of Language” by Guy Deutscher. It explains how the complicated structures in modern languages were created organically without central planning. It gives plenty of concrete examples, but is for laymen not professional linguists.

    A side effect of reading it is it makes one more relaxed about the things you’re complaining about.

    • Thanks for the suggestion, it sounds like a book I might enjoy. I did like Pinker’s The Language Instinct (but do not recommend How the Mind Works).

      I know language evolves, I know words mean whatever people think they mean, and I know a construction that is wrong in one era can be fine in another. I have no problem with any of that.

      But part of the way language evolves involves people resisting usages they don’t like. For instance, many of us object to ‘dominate’ as an adjective, as in ‘The Patriots are a dominate team.’ Dare I say all right-thinking people object to this? I do dare. The fact that people say this and I know what they mean, yeah, so what, it’s still wrong. I’m all for new usages when they improve something about the language, and there are plenty of those, but the use of ‘after’ that I find objectionable is not one. I know I’ll be the last one defending this hill, and that’s ok, somebody has to be the last one.

  10. My guess would be it is a rule of thumb that people figured out games a readability index of some kind. I tried:

    16-year-old bicyclist seriously injured after being hit by car in Norfolk

    Vs

    16-year-old bicyclist seriously injured by being hit by car in Norfolk

    https://readable.com/text/

    The “by” version has a lower grade level, and better “reading ease” score though. That amounts to one less “readability and quality issue” according to that site. So I dunno, maybe it is a different index? Or maybe I need to try out more examples?

    • “injured by being” sounds REALLY AWKWARD to me, and this may be the reason for the change.

      Norfolk traffic accident seriously injures 16-year-old

      sounds great, but it puts the emphasis on the location…

      If you want to emphasize the 16-year-old I guess something like:

      16-year-old injured in Norfolk traffic accident

      • How about 16-year-old bicyclist seriously injured when hit by car in Norfolk. Just ‘when’ instead of ‘after’.

        Or ’16-year-old bicyclist hit, injured by car in Norfolk’

        There are plenty of workable constructions.

        • How about 16-year-old bicyclist seriously injured when hit by car in Norfolk.

          This gets a ReadablePro rating of B, the original two got A’s.

        • It gave ’16-year-old bicyclist seriously injured after being hit by car in Norfolk’ an A, but ’16-year-old bicyclist seriously injured when hit by car in Norfolk’ a B? WTF? Et tu, readable.com?

        • I changed it to (there is no typo here):

          16-year-old bicyclist seriously injured when hit by car in Norfolk a

          And then it returned an “A”.

  11. My bugbear is “must” vs “should” ; in the English I grew up with
    these were usefully different.

    “You must not rob banks or you will go to jail”
    but now we get
    “Assad must go” (NYT)
    “The Federal Reserve must stand up to Trump” (NYT)

    almost always they mean “Assad should go” etc.

    But language does change, and I’m getting old…

    • I see your point. “Assad should go” seems pretty wimpy though. “You should go see Marriage of Figaro, and Assad should go too.”

      Perhaps there’s an implied “…or else” that makes these work. Assad must go or we will forcibly remove him. The Federal Reserve must stand up to Trump or it will lose all credibility.

      I’m not going to join you on your hill, I’ve got my own hill to fight on. Good luck though.

    • Both “must” and “should” have been replaced by “needs to.” “You need to stop robbing banks.” (“Assad needs to go,” is a bit ambiguous.)

  12. I think the real answer is that the journalists have a practice of wording that avoids libel suits. They get educated like this: Journalist: “Driver crashes into bicyclist.” Lawyer: “Do you actually know that the driver caused the crash.” Journalist: “Bicyclist seriously injured by car.” Lawyer: “Better, but you still can confirm that it was the driver’s fault.” Journal: “Bicyclist seriously injured in car crash?” Lawyer: “Look, have you spoken to the doctors, do you know which injuries were the result of the crash? Maybe he had a heart attack while ridding his bike.” Journalist: “Bicyclist in serious condition after crash with car?” Lawyer: “My work is done, and the English language is the worse for it.”

  13. The choices reflect how we imagine. The shift to ‘after’ reflects a construction that there are different divisible and describable, like a truck and a bicyclist, and that after these came together the bicyclist was in a state of being injured. A construction like in an accident is a different perspective, one that focuses on the accident as the nexus of the image, so it brings in ideas of what happened and why. A construction of by a truck describes a process, but of course at a level without detail so you imagine. My point: each one of these choices presents a different perspective which orders thoughts. One isn’t wrong versus another, but one may not fit as well as another, particularly when priors are involved and thus you are looking for a fit that already exists in an idealized form in your head.

  14. I checked Nexis-Uni for “injured after” in news articles in the three months Aug. 1 – Nov. 2 in three years.

    2019: >3000
    1999: about 1000
    1979; 55

    • You can only say killed by, or dead after, but injured serves both roles. I don’t know how to put this more formally, but that’s what has gone wrong here.

        • Yes, you’re right, Justin, that ‘Teen injured after car accident’ could mean ‘Teen is [still] injured after car accident’. I shouldn’t have used ‘injured’ in my example.

          ‘Teen killed after car accident’, though, that’s a lot harder to defend (although Daniel and others do defend it!) I think ‘Teen dead after car accident is OK’, though it still bugs me because I’d rather let the causality show. But ‘Teen killed after car accident’…as I said above, that only works for me if someone strangled her in the ambulance on the way to the hospital.

          All of this was said earlier in the comment thread, but it’s certainly excusable that you didn’t read through all of those dozens of comments.

        • Phil:

          I doubt many people are reading this thread anymore, but it strikes me that the arguments that the linguists are making on this thread are “proving too much,” as it were. I think they’re making a generic linguist argument against prescriptivism. And the right response on your part is not that the “injured after” formulation is wrong, or even factually incorrect, but that it’s misleading because it has an implication, perhaps only a soft implication, of a different sequence of events than actually happened.

          Just by analogy, I’ve complained, and continue to complain, about the use of terms such as “working class” and “blue collar” in political discourse. My problem is that the terms are vague: on one hand they are often used to refer to subgroups defined based on income or education level, but on the other hand they often evoke the image of white men. So I think these terms are misleading. I’m not arguing prescriptively that they are “bad English”—yes, I know about linguistic drift!—rather, I’m arguing that in the current context they can lead to confusion or even political bias. I think you’re making a similar argument regarding “injured after,” but the blog discussion got derailed into a general discussion of prescriptivist vs. descriptivist linguistics, where the linguists are thinking that you’re simply misunderstanding Linguistics 101.

Leave a Reply to Keith O'Rourke Cancel reply

Your email address will not be published. Required fields are marked *