Grater, grader, whatever

« previous post | next post »

Dan Hanzus, "Gruden on DJax: 'He wants to block, he just is little'", Around the NFL 10/16/2014:

The Washington Redskins did not sign DeSean Jackson to be a road-grater.

The 5-foot-10, 178-pound wide receiver gets paid to make big plays, not clear the way for them. But that didn't stop several D.C.-area media members from calling Jackson out for his apparent lack of effort in blocking situations during Sunday's loss to the Arizona Cardinals.

Perhaps the loudest voice came from former Redskins tight end Chris Cooley, who offered up a stinging critique of Jackson's game after some film study.

"Do not allow number 11 to ever be involved in blocking for screens, blocking for bubbles, picking for players in the pass game, (or) run plays to his side of the line of scrimmage," Cooley said, via the DC Sports Bog. "He WILL NOT TRY on them. Do not put him in those situations."

These are road graders. And these are road graters, though I think you get a better selection if you ask for pictures of road scarifiers, which are actually a thing.

Of course, "grader" and "grater" are homophonous for most Americans. And I guess that some people are think of grating as having something to do with shaving stuff off of a surface and thus making it flatter, so that they reckon that a big machine for making road surfaces flat could plausibly be called a "grater". And maybe these same people are not familiar with sense of the verb grade that means "To level or smooth to a desired or horizontal gradient". So, road grater.

And Mr. Hanzus is far from the first person to go this way. He's not even the first person to make it into Language Log for using this eggcorn.

[h/t Glenn Bingham]



  1. Gregory Kusnick said,

    October 16, 2014 @ 8:27 pm

    So Jackson's grade as grater is "not the greatest".

  2. goofy said,

    October 16, 2014 @ 9:26 pm

    Are they homophonous? I was under the impression that the vowel was longer before phonemically voiced consonants?

  3. Rod Johnson said,

    October 16, 2014 @ 9:42 pm

    That's probably true, but in American English at least, the t/d distinction is neutralized to some sort of tap or flap, which is voiced. So if there's lengthening, it happens in both words.

  4. goofy said,

    October 16, 2014 @ 9:57 pm

    Yes the flap is voiced, but the vowel is longer before *phonemically voiced* consonants. So it is longer in "grader" than in "grater".

  5. Nathan said,

    October 16, 2014 @ 11:00 pm

    I would need to see spectrographic evidence of a difference in vowels. I'm convinced they're 100% homophonous.

    [(myl) At least for most American speakers — and for the vowel in grade/grate — you're right.]

  6. Bob Ladd said,

    October 17, 2014 @ 1:17 am

    @Nathan,myl: On the basis of some informal experimentation some years ago, I wouldn't be quite so categorical. I suspect that if you looked at a large enough set of natural productions of grater (and rater, waiter, traitor, etc.) and of grader, raider, wader, trader, you would find a small but significant difference in mean duration and/or vowel quality. Most individual tokens would not be reliably identifiable as containing T or D, but there would still be a difference in the mean. That is, I think this is probably a case of "incomplete neutralisation" of the sort that has been found for German final devoicing and Dutch voicing assimilation. This explains both the intuition that they are homophonous and the intuition that there's some subtle difference. What it says about the nature of phonemic contrast is a different topic.

    [(myl) My own "informal experimentation some years ago" suggested the opposite conclusion, though of course we would need unattainable statistical power to prove that there's no difference at all in the means. It's probably worth doing a more careful experiment — though a laboratory study involving minimal pairs in matched contexts is exactly the wrong thing to do (though such studies have been published in some other incomplete-neutralization cases), since it evokes facultative disambiguation of an unnatural kind. And studies based on natural speech have got a whole lot of other problems. So it's not easy…

    Let's say at least that perceptual classification is effectively indistinguishable from chance, for tokens taken from non-facultative contexts.

    Of course there are some varieties of American English (including mine) in which "writer" and "rider" have completely different vowel qualities. So any generalizations need to be particularized to the vowel as well as the dialect/idiolect.]

  7. Ellen K. said,

    October 17, 2014 @ 9:03 am

    I hear a distinct length difference between grade and grate/great when I say them to myself. (Whether this comes out reliably in actual speech is another matter.) But not between grader and grater/greater.

    And, looking for a pair or words with a different consonant, same thing. Leg is longer than lake. But legger and laker are both short. (Yeah, that one only works for some of us, those who have the same vowel in lake and leg.)

  8. Jerry Friedman said,

    October 17, 2014 @ 12:39 pm

    Is there a name for nicknames such as DJax and J.Lo? Can I say "DJax is his hip-hop name" or something like that?

  9. popegrutch said,

    October 17, 2014 @ 1:43 pm

    So, is this just a Google glitch, or did you get identical results for set 1 and set 2 also? Since both words evidently refer to exactly the same thing, maybe there's no point in making a distinction in spelling? (Links to Google searches are problematic at best, since Google tailors results to the user's history. They may just figure I don't care about the difference, but you do.)

  10. chh said,

    October 17, 2014 @ 3:22 pm


    When you search a term or phrase that looks a misspelling, Google automatically corrects it, returns results for the corrected search, and asks you if you want to search for what you actually typed.

    If you click "Search instead for road grater" you'll get a different set of results.

  11. J. W. Brewer said,

    October 17, 2014 @ 3:41 pm

    To take an example going the other direction, "cheese grader" is both an eggcorn for "cheese grater" but also a licensed profession in, at least, the State of Wisconsin. I'm not sure what the penalty is for purporting to grade cheese in Wisconsin without a license, but it's probably better to pay the $75 for the license and not have to find out.

  12. Aaron Braver said,

    October 17, 2014 @ 6:28 pm

    I've had a look at this in a paper coming out in Lingua. For nonce words, on average (New Jersey) speakers of American English have pre-/d/-flap vowels 5.69ms longer than pre-/t/-flap vowels. Just as Bob Ladd suggests above, this incompletely neutralized difference appears to not be perceptible.

  13. goofy said,

    October 17, 2014 @ 7:13 pm

    What about non flap environments like "grate/grade"? The vowel is definitely longer in "grade" right?

  14. maidhc said,

    October 18, 2014 @ 3:17 am

    As someone with a basically north central North American accent, I find myself degrading t to d quite frequently. Since I spend a lot of time talking to people whose native language is not English, I try to consciously emphasize the difference so as not to confuse them, but it is a challenge for me.

    (One of several things I try to work on in my own speech. I've had people tell me they have difficulty distinguishing between "can" and "can't" in North American speech, so that's another one I try to bring out.)

  15. chh said,

    October 18, 2014 @ 10:56 am

    One thing that I wonder about in these kinds of discussions about incomplete neutralization is the learning of the underlying status of pairs like "utter/udder".

    If speakers know that "utter" has a /t/ and "udder" has a /d/ (Aaron's paper cites a priming experiment that suggests they do), but they are unable to perceive the acoustic difference between the two, how did they learn the correct underlying phoneme for the words in those pairs in the first place?

    It would be nice to see whether and how often people learn the wrong underlying forms for those words.

  16. A. C. said,

    October 18, 2014 @ 5:50 pm

    I suppose for a language instructor, grading can be grating at times.

  17. Pflaumbaum said,

    October 18, 2014 @ 7:22 pm


    I was wondering the same thing. If it really is imperceptible – not just consciously imperceptible – it's hard to see how it could be passed systematically to the next generation.

  18. Pflaumbaum said,

    October 18, 2014 @ 7:24 pm


    I was wondering the same thing. If it really is imperceptible it's hard to see how it could be passed systematically to the next generation.

    Is the idea that it is perceived by the brain but not made available for conscious recollection/analysis?

  19. dw said,

    October 19, 2014 @ 1:51 pm

    Excellent road. A+

  20. Bloix said,

    October 20, 2014 @ 3:41 pm

    I have a sense the American lack of a t/d distinction is the second most important difference between most American and most English accents (after "r"). I suspect that although we don't consciously notice it, Americans hear the English distinction as fussy, while the English hear our "flap" as sloppy.

RSS feed for comments on this post