I heard that word pronounced “nego-see-ation” for the millionth time today (this time by an NPR newsreader) and I finally cracked. Yes, my descriptivist faith tells me if so many people say it it’s not “wrong,” it’s just an alternate pronunciation (in fact, I just checked Merriam-Webster’s latest Collegiate and I see it’s listed as an alternate), but I’ve hated it all my life and I just have to post about it and get it off my chest. I used to think it revealed a prissy fear of sounding colloquial by using the /sh/ sound, but now I’m wondering if it’s good old dissimilation, the -sh…sh- sequence producing -s…sh-. If so, I guess I’m slightly reconciled.
This has been the Languagehat Venting Hour, or rather Minute. Our regular programming will resume tomorrow. Thank you for your attention.


  1. My wife says “nursery” with two syllables and an /sh/ sound in the middle. For some reason, it makes me cringe just a little. There’s no good reason for my reaction; I pronounce “grocery” the same way. But I say “nursery” with three syllables and an /s/. Maybe it’s the /rshr/ sandwich in the middle that I don’t like.

  2. FWIW I am inclined to agree with your first view, that the /s/ pronunciation is a marker of “formal register”. I notice newsreaders here (Victoria, Australia) also like to say “controver-see-al”, where I would say “controvershal” (excuse my pseudo-phonetic renderings). In this case I suppose you could argue that they are bringing the pronunciation into line with the noun “controversy” for the sake of morphological transparency. But I think it’s just hypercorrection.

  3. Jonathon: What the hell? You say “grosh’ry”? Is that common?

  4. strangeguitars says

    nomis: “Groshry” is very common in American English. If you have a word with at least three syllables, and one of the unaccented middle syllables has the “er” sound, and the syllable after that starts with a vowel sound, the “er” sound will merge with the following syllable.
    Examples: every is pronounced ev-ry, restaurant is pronounced rest-rahnt, different is pronounced diff-rent. So grocery would be pronounced grow-sree.
    But English doesn’t have any words or syllables that start with “sr”, so people might unconsciously find that strange. There are a lot of words and syllables starting with “shr”, so they might use that instead, making grow-shree.

  5. Does anyone have an explanation for why the normal pronunciation of “equation” has a “zh” sound rather than “sh” like other “-tion” words? I actually do use the “sh” version, but only when it means the act of equating rather than a mathematical expression.

  6. And don’t the people who say “nego-see-ation” also say “nego-see-ate”? If so, it’s not dissimilation. What about “ini-see-ation” and “appre-see-ation” and “asso-see-ation” and “pronun-see-ation”? I do say the last, and I’m not sure about the third.

  7. I bet it is a kind of dissimilation, partly driven by an attempt to clearly enunciate the “i” — “negoshiashun” becomes “negoshashon” (“negoshate”, etc.) quite easily if you’re speaking fast, and there’s the added danger of a say-it-don’t-spray-it disaster for some. I also know people who say “appresiate” etc. and I’ve always assumed (there’s another one) this was why.
    Somewhere in the past 5 or 10 years I picked up the habit of saying “fid’ty” instead of “fifty” for similar reasons.

  8. This horrendous mispronunciation of “negotiate” began as a mistake during one frantic report from the scene by a BBC reporter during the Falklands war. Suddenly every American twit in front of a camera reading news off a teleprompter thought he’d sound more edumacated if he said it that way. It’s been like a burr on a wool sweater ever since.

  9. I (Australian) used to say “groshries”. For some reason I’ve unconsciously slipped into “grosseries”. Spelling pronunciation?

  10. strangeguitars: Interesting, thanks. I’m pretty sure I’ve never heard “groshry” before (I’m from NZ), although of course we do elide unstressed vowels. I’ll be listening out for it now.

  11. ‘Association’ and ‘pronunciation’ both normally have the /s/ sound, KCinDC, at least in my American dialect. ‘Pronunciation’ would sound especially weird with a /sh/ there- I would assume the person had a lisp.

  12. I’ll quell desire to comment upon the US use of “alternate”.

  13. This person living and working in London negoshiates for gro-series in gros’ry stores; but his negosiashuns for plants are in nur-series rather than nur’sries.

  14. Well my guess is that groshry is a way of getting away from associations with “gross” – not a problem for UK speakers like me – for us gross means big and bad, normally (or a quantity), not disgusting.
    My own manic cringe came listening to a recording of Eliot reading Prufrock: in “I have measured out my life with coffee spoons”, “measured” is pronounced “mezzyured”. YUCK!

  15. Shurely shome mishtake?

  16. So will I, Sili. We don’t want to offend our host…
    On altered sibilants, though, I call to the stand a very famous Australian philosopher, who (I think I read) is our most cited academic tout court. He always says easier with a /zh/ (as in pleasure). So much is this ingrained for him that he nearly always says easy with a /zh/, also. I find it endearing, and Sean-Conneric.
    This business with syure (along with assyurance, etc.), syuit, and the rest is horribilis auditu, I agree. What about nature, though? Is the /tch/ replaced, here and there? We think also again of Pound, and kulchur.
    I’d be interested to hear something about the incidence of strained-formal /bi’ond/ (‘=glottal stop), for beyond. Common enough in Australia, to the trained and attentive ear. Elsewhere?

  17. I can appre-see-ate your discomfort!

  18. Tom Wootton says

    The principle is sound, which is, despite your wilful descriptivism (shakes head sadly at such a bright young thing so sadly misguided), people who say iss-yoo and tiss-yoo are incredibly irritating and tend to wear bright clothes.

  19. Noetica: I’ve heard /bi’ond/ before. One of my friends says it that way.

  20. I had always assumed that the American tendency (and it seems more pronounced on the west coast than in the east) to pronounce ‘str’ as ‘shtr’ (so that ‘street’ is pronounced ‘shtreet’) was from German influences. Strangeguitars attributed the ‘sr’-to-‘shr’ transition as being due to the fact that no English words begin with ‘s-r’. There are plenty that begin with ‘s-t-r’, so his explanation doesn’t work here. I figured once you’re pronouncing ‘str’ as ‘shtr’, it’s only a hop and a skip (not even a jump) to changing ‘gros-ry’ into ‘grosh-ry’.

  21. What about the cockney pronunciation of “stupid” as “shchupid”?
    For that matter, is that really “shtreet”, or is it maybe “shchreet”, to go with pronouncing “truck” as “chruck” and “drum” and “jrum”?
    And if the pronunciation came from German influence, why would it be more common on the West Coast?

  22. I’m not sure where most of you come from, but I was raised in Boston where my grandmother taught us to speak in some sort of T.S.Eliot speak, most likely in the hopes that we would not be pahkin the cah.
    I do know that my ex-mother in law, who worked in the Home Office, had a peculiar way of pronouncing “schedule”. Sounded like “shed’-ule”. She seemed fairly normal other than this. Not even my grandmother or Eliot went this far.

  23. parvomagnus says

    The problem with the German explanation is no one says “shmile” or “shtone”. I’d blame the ‘r’ itself for all this. Keep in mind the English ‘r’ is rather a weird sound, and there are several different ways of making it; I don’t know about other varieties of English, but in America the retroflex variant is popular, which involves curling the tip of the tongue back. For people with this ‘r’, other sounds before the ‘r’ tend to assimilate to it; hence “chree” for “tree”.
    For me, at least, this is not an actual “ch” like in “cherry”, but a retroflex, aspirated ‘t’. Since this is also a weird sound, not otherwise in American English, that exists only in conjunction with the ‘r’, when people try to recreate it outside of that environment (“he says ‘tree’ with a ‘ch'”) I’d guess they’d use the independent sound most like it.
    Same for ‘street’; the tongue’s being preemptively curled back for the ‘r’, and this wreaks (what else?) havoc on the sounds before it. It may be some people actually pronounce these sounds as normal “ch”s and “sh”s, but then I’d guess those sounds passed through that stage.
    As for “negossiation”, the “fancy-pants” explanation might be right. Both “sh”s in that word were originally “s”s in middle English, and my distant, American impression of British English facts is that the old-style “sy” for “sh” sounds like high style, so I could see how that might be generalized to “‘s’ is better than ‘sh'”. Sort of like the pseudo-Greek “processeez”. I have no idea here though, just guessing.
    An admitted drawback to the type of analysis above is it distracts you from the truly important part of language analysis, which, as Tom Wootton pointed out, is mocking people you don’t like.
    Also, never heard “beyond” with a glottal stop. Is the “y” still pronounced?

  24. Should I point out the /str/ -> /skr/ shift in AAVE? (I sometimes notice it in my own speech). For those of you who aren’t familiar with the shift, ‘street,’ ‘string,’ ‘stream,’ and ‘strong,’ for example, are pronounced ‘skreet,’ ‘skring,’ ‘skream,’ and ‘skrong’.
    I would guess that this somehow mirrors the school-shul, sholde-skolde, shield-scild, sh-sk dynamic, bearing in mind the standard American ‘shtreet’ pronunciation Frank noticed.
    madame I, it seems like your ex-mother-in-law had a normal British pronunciation of ‘schedule’. This is from
    “Modern spelling is 15c., in imitation of L[atin]; the modern British pronunciation (“shed-yul”) is from Fr[ench] influence, while the U.S. pronunciation (“sked-yul”) is from the practice of Webster, and is based on the Greek original.”
    Apparently it was spelt/spelled ‘sedule’ or ‘cedule’ before that.

  25. mollymooly says

    I don’t associate the elision of unstressed -er- with America; at least, I consider a three-syllable pronunciation of “interest” to be markedly American. As is three-syllable “medicine”. And one-syllable “it’d”.
    Do Americans call that rapper Fiddy Cent or Fiffty Cent? In Ireland it’s always the latter.
    Irish newsreaders seem to have been ordered lately to pronounce issue as “iss-yoo” intead of “ish-oo”. Oddly, there’s been no crackdown on the ones who rhyme says with days.

  26. I had wondered about the /st/ > /sht/ thing before, and how it usually seemed to be Americans (?Californians) who said it that way. My completely unfounded guess was that it came about because of the collision of a laminal (flat of the tongue) /s/ and an apical (tip of the tongue) /t/. When I (BrE, RP) pronounce these words, the /t/ becomes laminal, but I wondered whether the assimilation could work the other way, giving an apical /s/ (like in Dutch), which sounds to others like a /sh/ sound.
    Of course this is pure speculation, so please feel free to shout me down….

  27. And don’t the people who say “nego-see-ation” also say “nego-see-ate”? If so, it’s not dissimilation.
    Excellent point. And I’m glad to see I’m not the only one bothered by this.

  28. The pronunciation of “street” as “shchreet” is an assimilation of the /s/ to the /t/, which has been palatalised because of the following /r/. The same thing happens, at least in NZE, with “stew”, which goes from “styoo” > “schoo” > “shchoo”.

  29. Ah, Madame, the British and Australian pronunciation of schedule is indeed with /sh/, not /sk/. And schism is with /s/, not /sh/. But the young now have /sh/ in both.
    Parvomagnus, in beyond when it is pronounced /be’ond/ there is nothing corresponding to the “y”. It is replaced by a glottal stop, when people unused to formal situations hypercorrect to avoid anything shamefully intervocalic that will give them away as perfectly competent natural speakers of their mother tongue. I am interested that Justin has heard it. I wonder if his friend does other things like that: over-use of thus (or even thusly); if-less conditionals (Had I the time, I would explain); whilst for while?

  30. I meant: And schism is with /s/, not /sh/ or /sk/.

  31. My mother used to pronounce shrimp as srimp, which I take to be a rather rare variant, but not unique. Not sure how she came by it, having come of age in the Shenandoah Valley. Sort of goes against the grain of str > shtr, which BTW is also characteristic of Hawaiian English.

  32. Noetica, I’m surprised to hear that Australians pronounce schedule as /sh/ rather than /sk/. I’ll have to start listening for it. In NZ it is almost universally /sk/.

  33. Nomis, I grew up in Australia, and “skedule” is how and all my friends pronounced it. The “shedule” pronunciation, for me, has the same mildly silly, unsuccessful-attempt-at-fancy-pantsery ring as “an history lesson” with the ‘h’ pronounced.

  34. (Uh, no offence, Noetica…)

  35. parvomagnus says

    Noetica: That’s what I figured with ‘beyond’. Is it just that word, or is it some yod-dropping type stuff?
    I sometimes use fronted subjunctive-relics in place of ‘if’ myself, for “be”, “have”, and modals. Never for “do” though; stuff like “did their Catullus walk that way” has always struck me as bizarre. Aside from that, I’d never thought of it as particularly British before.
    The “shchreet” stuff’s still I think a slightly different case than “shchew”. I’m unware of any dialect of American English with a palatal ‘r’; Of the two main allophones, the alveolar approximant is articulated alveolarly, just like the /t/s and /s/s, and the retroflex approximant is, to quote wikipedia, “postalveolar without being palatalized”. So I’d guess the assimilation is to the retroflex part. I’m not sure what the distribution of the two variants is, but I think they’re both pretty wide spread.

  36. Ha ha, Matt. No offence taken! I said: But the young now have /sh/ in both. But in fact, I should have said that they say /sk/ in both. Dammit, I made a mess of that post. You may have grown up in Australia: so did I, but I think I must have done so before you did. Schedule is still “corrected” to /sh/ in Australia, but the Americans have long ago won the tongues of the young.

  37. As for an history lesson, that is more archaic than the much more common an historic occasion. The acceptability of this an depends on the stressing of the first syllable of the following h-word. Not just whether it bears the main stress or not: if it is stressed less than the second syllable, the an is more acceptable. So an historical account is more likely than an hysterectomy. That’s one consideration; another has to do with mere idiom, which works for an histor- in general, and against such nonce-formations as an hyster- and an humiliating defeat.

  38. Matt, Noetica: the Oz situation is the same as NZ then. Only conservative speakers use /sh/ … although I should point out that I say it that way (my mother is a prescriptivist you see) and sure enough, anyone of my generation (30-35) mocks me mercilessly. But I enjoy my moments of half-hearted fancy-pantsery. I also use fronted subjunctives (e.g. “Had I known that, I never would have …”) but only in writing.

  39. parvomagnus: I wasn’t suggesting that /r/ is palatal. In fact, the change of /t/ to /ch/ when preceding /r/ is not really palatisation at all (more like post-alveolo-affricativisation), and as you say is probably triggered by the retroflex feature of /r/.
    I would agree that the changes to the /t/’s in ‘street’ and ‘stew’ are different things, triggered by following /r/ and /j/ (palatal glide) respectively, but the subsequent change of the preceding /s/ is the same in both cases, since both /t/’s end up as /ch/.
    I hope that makes some kind of sense to someone.

  40. There’s a sign marking a neighborhood in my town that declares it “An Historic Community.” Granted, the ‘h’ syllable is unstressed, but it still makes me want to egg their houses and/or curse the whole neighborhood.
    Good example Joel! I find it very difficult to produce /shrimp/; it sounds like /skrimp(s)/ instead. I’d agree that /srimp/ is pretty rare, though.

  41. Great discussion! I’m glad you all managed to amuse yourselves while I’ve been offline. (Router problems, don’t ask.) Anyway, I’m back and glad of all the enlightening (if sometimes confusing) comments!

  42. Anyway, I’m back and glad of all the enlightening (if sometimes confusing) comments!
    In the Rhode Island dialect (referred to locally as Vo Dilunese) confuse is pronounced cunFOOS, which I find charming.

  43. Sounds like they’re trying to sound fancy. Or British, since it seems more “proper,” even if the word isn’t *supposed* to be pronounced that way.

  44. Loan words usually become nativized in a language sooner or later. Even the French have turned Spanish mosquito into moustique and German Sauerkraut into choucrout.
    British English “Nego-see-ay-sun” reflects more the original French pronunciation of the word just like the British pronunciations of the words picture and vase: “pik-tyoor” and “vahz.”
    However, most Americans say “nego-shee-ayshun,” pik-cher,” and “vayz” which are actually a little more Anglo-Saxon than the British pronunciations.

  45. But it’s not “British English ‘Nego-see-ay-sun'”—the traditional British pronunciation is with -sh-, just like the American. It’s a new development, due (apparently) to fear of sounding unposh.

  46. David Marjanović says

    My completely unfounded guess was that it came about because of the collision of a laminal (flat of the tongue) /s/ and an apical (tip of the tongue) /t/. When I (BrE, RP) pronounce these words, the /t/ becomes laminal

    So I can stop trying to curl my tongue till it hurts. 馃檪 Being a native speaker of southern (!) German, my alveolars are all laminal; I was only told a few years ago that the English ones are apical — I had noticed something sounded wrong about my /d t n/ when I spoke English, but I hadn’t figured out what. I seem to have mastered that by now, but I find an apical [s] much more difficult to get right. So I can stop trying. 馃檪 馃檪 馃檪

    Good example Joel! I find it very difficult to produce /shrimp/; it sounds like /skrimp(s)/ instead.

    Looks like your /r/ is always retroflex. With an alveolar one, /Sr/ is much easier than /sr/, IMHO.

    and German Sauerkraut into choucrout.

    Choucroute. Otherwise the t would be silent.
    (BTW, what’s going on in the first syllable of this word? Analogy to chou-fleur?)

    the British pronunciations of the words picture and vase: “pik-tyoor”

    I was taught British English at school, and I wasn’t taught that pronunciation for picture. It was exclusively reserved for mature.

  47. That “chou” is simply French for “cabbage,” no? Larousse Etymologique derives “choucro没te” and “sauerkraut” both from Alsacien s没rkr没t, meaning herbe sure, which in fact doesn’t mean what we call sauerkraut at all, but what our kids call “sourgrass,” a kind of sorrel. Chou=cabbage; cro没te=crust; I recall sauerkraut in the making as shredded cabbage with a crust of salt.

  48. Well, if choucro没te is from Alsacien s没rkr没t, then the chou isn’t from the French for “cabbage,” it’s just been remodeled to look like it.
    the British pronunciations of the words picture and vase: “pik-tyoor”
    No, picture is traditionally PIK-che(r)—Daniel Jones gives no other pronunciation. “PIK-tyoor” is a prissification.

  49. Most of us in the UK say pik-cher. A few prissy arty types say pic-tyoor, but not many. Languagehat, everyone deserves a good rant every now and again. Enjoy the moment!

  50. I don’t think my /r/ is always retroflex, David, but that I learned the word with the /k/ and so have to concentrate to avoid using it. I have alveolar /r/ in other places, but it was a valid guess.

  51. David Marjanović says

    Sauerkraut is acidic because the fermentation produces lactic acid, and named after that fact(sauer “acidic”). No?
    (Same word as English sour. The point of intersection is sour milk — saure Milch, which is both rotten and acidic.)

  52. i’m gonna go watch a filum.

  53. Charles Shere: Chou=cabbage; cro没te=crust; I recall sauerkraut in the making as shredded cabbage with a crust of salt.
    Have you actually seen this? or just heard about it?
    LH: Well, if choucro没te is from Alsacien s没rkr没t, then the chou isn’t from the French for “cabbage,” it’s just been remodeled to look like it.
    LH is right. CS’s derivation is a case of folk etymology.
    The French name of this food is choucroute, without the circumflex which exists in cro没te meaning ‘crust’. It is a traditional Alsatian preparation which is of course eaten throughout France nowadays but is not indigenous to other French regions. Also, in the days before refrigeration salt was used to preserve a variety of foods, none of which have the word cro没te in their names, and choucroute as bought and consumed by the average French speaker does not come encrusted with salt. Furthermore, the word cro没te without a qualifier or context calls to mind a hard crust of bread or pastry, not of salt.
    For a purely French name of the preparation emphasizing the role of salt one might conceivably expect something like chou sal茅 ‘salt cabbage’ or chou au sel ‘cabbage with salt’, but a compound chou + cro没te does not seem well-formed (unless it was for some reason the name of a specific variety of cabbage), and this is typical of adaptations of foreign words such as Alsatian s没rkr没t.
    An example of such an adaptation in English is mushroom which is not from mush + room but from the Old French ancestor of the word mousseron, a type of mushroom.

  54. John Cowan says

    Specifically St. George’s mushroom, Calocybe gambosa, which first appears in the UK on St. George’s Day, April 23. This mushroom grows in moss, the Old French for which was mo(u)sse, hence the name mousseron. Nowadays,mousse in French has shifted semantically to ‘foam, froth’, and the culinary mousse is an extension of this meaning. The PIE root is pretty widespread, and it’s not clear if the OF form is a Germanic borrowing or native or a semantic mixture (as the OED thinks). In Germanic the meaning was extended to ‘bog’, but this meaning has been lost from Standard English.

  55. David Marjanovi膰 says

    In Germanic the meaning was extended to 鈥榖og鈥, but this meaning has been lost from Standard English.

    Are you saying moor is a Verner doublet?

  56. John Cowan says

    No, moor is the o-grade of mere ‘pond’, where the /r/ is in the root (cf. Latin mare, Russian more). The Verner doublet, if that’s what it is, is mire, which however is a borrowing from Old Norse m媒rr. However, Wikt says the two were already distinct at the PGmc level: *mus膮 > moss vs. *miuzij艒 > mire.

  57. I too have heard the word “negotiation” pronounced many times with the s sound. It’s usually used by high diplomats and the like and “highly educated” people. It does sound more educated and I think if I tried to use it it wouldn’t come out right, even though I believe I’m educated as to formal schooling and life experience.

Speak Your Mind