« previous post | next post »

The current disaster in Japan raises the question of the origin of the word tsunami. It is from Japanese 津波, where 波 [nami] is "wave" and 津 [tsu] here means "harbor". It was apparently first used in English in 1897 by Lafcadio Hearn in his Gleanings from Buddha Fields. The Japanese Wikipedia article contains a discussion of early English usage.

In English the word is pronounced [sunami] rather than [tsunami] since English does not allow syllable-initial [ts]. This is yet another example of insane English spelling practices and of the fact that they cannot be blamed entirely on the preservation of archaic spellings. The word could perfectly well have been borrowed into English as sunami. The person learning to write English must memorize the fact that this [s] is written <ts> for no reason at all. Note that the English spelling does not even have the virtue, whatever that might be, of preserving the Japanese spelling since Japanese is not written in Roman letters.



  1. Martin J Ball said,

    March 11, 2011 @ 1:25 pm

    The argument that the 'ts' spelling is "for no reason at all" is surely a tad strong.
    The reason is clearly to reflect the original pronunciation even if the majority of English speakers avoid an affricate. It also means that it can be recognized in other languages (e.g. German) which do use a similar realization to the Japanese.

  2. RW said,

    March 11, 2011 @ 1:26 pm

    I am surprised that you say it is pronounced [sunami] in English – I always pronounce the initial t, and in my experience most English speakers do as well.

  3. Keith said,

    March 11, 2011 @ 1:26 pm

    Plenty of English people pronounce [ts] at the beginning of tsunami, tsetse, and tswana…

    It's true, maybe, that the consonant cluster [ts] does not appear at the beginning of any native English words, but we can pronounce it in words borrowed from other languages.


  4. Andrew said,

    March 11, 2011 @ 1:28 pm

    I heard someone on the Today Programme on BBC Radio 4 this morning pronounce tsunami with the initial "ts". So maybe they were trying to pronounce it the Japanese way for a disaster in Japan.

  5. Leonardo Boiko said,

    March 11, 2011 @ 1:36 pm

    Japanese is certainly (also) written with Roman letters. Not primarily, sure, but one does find rōmaji (Latin/Roman alphabet) writing occasionally in advertising, signs, clothing etc.— not to mention Japanese words cited in foreign-language texts like this one. There are a few minority organizations that support wholesale adoption of the Latin alphabet, there’s been literature written in rōmaji, and today there are two well-established orthographies (Hepburn and Kunrei), the latter one being taught on schools. A variation of these is currently the primary method of inputing Japanese texts on computer keyboards (“word processor rōmaji” or “input method rōmaji”) —though the output is converted to the usual kana/kanji mixture.

    The English spelling of “tsunami” preserves the Hepburn form, FWIW, which is the most intuitive to non-Japanese speakers (Anglophones might not pronounce the [ts] but the Kunrei form, “tunami”, would be even more misleading).

  6. carla said,

    March 11, 2011 @ 1:40 pm

    I agree with RW that your assertion that "In English the word is pronounced [sunami] rather than [tsunami]…." is somewhat overbroad.

    In my idiolect, it's [tsunami], and while I can't say for sure how other people pronounce it (I will have to start listening now), it's certainly unremarkable to me to hear [tsunami] – it wouldn't sound like an affected foreign pronunciation to me.

    I am a native speaker of US English, raised in NYC, now living in New England.

  7. Mark Mandel said,

    March 11, 2011 @ 1:44 pm

    I agree with Carla 100%, except that unlike her I am now living in Philadelphia.

  8. Russell Borogove said,

    March 11, 2011 @ 1:45 pm

    The English spelling does have the virtue of preserving the Japanese pronunciation, even if some English-speakers don't honor it.

  9. Jonathan Lundell said,

    March 11, 2011 @ 1:48 pm

    The Oxford American gives (t)suˈnɑmi. I pronounce the t myself, and that's what I hear, but I haven't been paying close attention. I just listened to the NPR hourly news, and it was hard to tell whether the NPR announcer had the t; a British reporter definitely did.

  10. Bill Poser said,

    March 11, 2011 @ 1:51 pm

    Sure, some English speakers pronounce initial [ts]. It is conceivable that this will spread to the point that initial [ts] will become a natural part of the English sound system. However, I strongly suspect that this is an elite phenomenon: those who produce [ts] are a small minority familiar with other languages and of cosmopolitan orientation.

    I am well aware that Japanese is occasionally written in romanization. It is, however, virtually never written in romanization for ordinary purposes. Input methods based on romanization nonetheless result in non-Roman text. The great bulk of romanized Japanese that one sees is in brand names and the like. It is exceedingly rare for a book, newspaper or magazine article, blog post, letter, email message, shopping list, diary entry, or any other ordinary Japanese text to be in romanization. As I pointed out in a previous post, even the Japanese Romanization Society does not produce its newsletter entirely in romanization.

    In any case, what is the virtue of making English spelling conform better to the spelling either of Japanese (even if it were routinely written in romanization) or German or other languages? I submit that in cases such as these we add to the difficulty of learning to write English, and of remembering how to do so, for a negligible advantage. How many English speakers are going to read something in German and understand the word tsunami because they know it as such in English who would not recognize it if English wrote it as it is usually pronounced in English?

  11. Leonardo Boiko said,

    March 11, 2011 @ 1:55 pm

    In Bill’s favor, reddit is currently filled with “too tsunami” jokes (from “too soon”. Being a non-native, I failed to understand the wordplay until reading this post.

  12. Thomas Westgard said,

    March 11, 2011 @ 1:56 pm

    Your description of English-speaking people not pronouncing the initial stop in "tsunami" didn't ring true to me, so I went to YouTube and watched about a dozen videos. What I found there was that people seem to begin the word by placing their mouths into the t-stop shape, and then starting the "s" sound from that initial position. It is a different sound from other, natively English words like so, sound, and soup. So I guess maybe you could define what you mean by not pronouncing the "t" a little more closely. There does appear to be some physical effort expended by the native English speakers that changes the sound, to my ear at least.

  13. Jonny Morris said,

    March 11, 2011 @ 1:58 pm

    The BBC pronunciation guide for presenters gives [ts]unami.

  14. Jean-Sébastien Girard said,

    March 11, 2011 @ 2:02 pm

    Says to me English word adaptation practice are weird. French does not not have /ts/ as an originally allowed cluster, period, and we still managed to pronounce that word fine. Same with the /p/s in words like psychology and pterodactyl (took me YEARS to actually acquire that purportedly correct pronunciation).

  15. Xmun said,

    March 11, 2011 @ 2:03 pm

    Here in New Zealand I have only ever heard "tsunami" pronounced with initial [ts].

  16. Xmun said,

    March 11, 2011 @ 2:04 pm

    Or /ts/ perhaps I should have written.

  17. Xmun said,

    March 11, 2011 @ 2:06 pm

    And add tsar to tsetse and tswana.

  18. Johanne D said,

    March 11, 2011 @ 2:07 pm

    Jean-Sébastien : je n'ai aucune difficulté non plus à prononcer Gbagbo, et toi?

  19. Leonardo Boiko said,

    March 11, 2011 @ 2:15 pm

    @Johanne: Gbagbo is easy next to Eyjafjallajökull.

  20. Outis said,

    March 11, 2011 @ 2:21 pm

    I have always pronounced tsunami with the [ts], and that was before I learned and cared anything about the Japanese language, and I heard it pronounced either way during the South Asian earthquakes. I strongly disagree that pronouncing [ts] shows any kind of cultural-linguistic sensitivity toward the Japanese. The initial [ts] may be foreign to English, but it is not entirely awkward. Many English speakers have the instinct to ignore the t, but many others will simply read it the way it's spelt.

    As for Japanese romanization, it is used much more widely than you suggest. Japanese people readily switch to writing in romaji when Japanese isn't available — which was the case for early text messaging and for Japanese living abroad.

    Lastly, English spelling of tsunami isn't just conforming to a Japanese standard, it is in fact the best phonetic representation of the Japanese word in English. It certainly preserves the etymology better than sunami. In any case, English spelling system does not conform to any consistent standard anyway, spelling it as sunami would be even more arbitrary than tsunami.

  21. James C. said,

    March 11, 2011 @ 2:29 pm

    The use of initial /ts/ is in nearly all English-speaking societies an affectation brought on by spelling and/or knowledge of Japanese. People who don’t know how to spell the word will produce it with an initial /s/. This is fairly easy to test, simply talk to less educated speakers when a tsunami is in the news. If they did not know the word before, they will quickly adapt it to an initial /s/.

    But in Hawaiʻi, where a significant portion of the population descend from Japanese immigrants, most people do pronounce the initial /ts/ rather than /s/. In undergrad linguistics courses many of the local students find it odd that other English speakers generally do not do so, and that initial /ts/ is phonotactically peculiar in English. The same goes for glottal stop /ʔ/, which is by no means marginal in Hawaiʻi due to its prevalence in Hawaiian placenames. Long vowels, present in Japanese, Okinawan, Hawaiian, and Samoan among other local languages, have not become quite as entrenched. My hypothesis is that this is because long vowels are a metrical phenomenon rather than a segmental one, but I haven’t investigated it seriously.

  22. Zythophile said,

    March 11, 2011 @ 2:31 pm

    That British English normally pronounces the word with initial "t" is confirmed by the poor-taste joke playing on the similarity of "tsunami" to "Toon Army" (the nickname of Newcastle United football club) that saw the footballer turned TV pundit Rodney Marsh sacked by Sky in 2005.

  23. Josh Treleaven said,

    March 11, 2011 @ 2:33 pm

    And I strongly suspect that using the word tsunami at all is an elite phenomenon. The vernacular is "tidal wave", inaccurate as it may be.

  24. Catanea said,

    March 11, 2011 @ 2:35 pm

    I was in the 1967 Alaska earthquake. I was a child. The radio warned us to move to higher ground away from the harbour because there was danger of a tsunami. So we did. I've never met anybody who didn't pronounce tsunami with a t. (Yet.) I'd've asked them about it.

  25. Matt McIrvin said,

    March 11, 2011 @ 2:38 pm

    I'm used to "tsunami" instead of "tidal wave" being an elite usage in itself (justified with appeals to factual consistency, since "tidal waves" have nothing to do with tides). But that may be an out-of-date impression.

    Google Ngram Viewer suggests that "tidal wave" was the standard English expression before about 1960 and "tsunami" has risen to approximately equal frequency since then. (But "tidal wave" hasn't dropped much; maybe it's still common in metaphoric use.)

  26. John Lawler said,

    March 11, 2011 @ 2:39 pm

    Simple test: speak, elicit, or record a phrase like "There's been a tsunami", making sure to say "a tsunami"; an initial vowel should make any [t] before the [s] stand out.
    It's interesting to note that, while English would describe the effect as a potential intrusive [t] before the [s] that's undoubtedly there in the word, in terms of Japanese phonemics the opposite is true; the [s] is an intrusive consonant that occurs after the [t] that's undoubtedly there in the word.

  27. Jerry Friedman said,

    March 11, 2011 @ 2:44 pm

    @Xmun: In my experience in America, "tsar" and especially "czar" are almost always pronounced with initial /z/.

  28. Ice Wolf said,

    March 11, 2011 @ 2:45 pm

    I can't recall much from High School in Ontario; though that was when I was first introduced to the word; but of my peers, it seems to be in free-variation between /ts/ and /s/. But more on the vowel quality on the following, I'm not that well at hearing the differences, but the syllable length seems to be constant, with more of a /tsu/ vs /su:/.

    To me anyhow, it seems to be in the similiar amount of variation between speakers that still retain the palatization of /u/ in words like Tuesday and News. (I hear both -u and -ju commonly enough that the "american" version doesn't sound nearly as foreign as it used to)

    Makes me want to ask people on the street and see if there is a correlation between the two.

  29. Josh Treleaven said,

    March 11, 2011 @ 2:49 pm

    Isn't it an interesting phenomenon how English likes to borrow foreign words that are just as wrong. We replace tidal wave, because it's not related to the tides, with harbor wave in a foreign language, even though it doesn't only affect harbors. Maybe we need a new word. How about "quakewave"? Let's all.

  30. Anie said,

    March 11, 2011 @ 2:49 pm

    I'm somewhat confused by the claim that the spelling "tsunami" in English doesn't reflect the Japanese spelling (since Japanese isn't written in Roman letters). Japanese does make use of two syllabaries, both of which distinguish tsu (つ) and su (す), and the English spelling "tsunami" accurately reflects that the word is written as つなみ in hiragana, not すなみ.

    (Of course, whether or not this is virtuous is another point entirely.)

  31. Eric P Smith said,

    March 11, 2011 @ 2:49 pm

    Although I am a native English speaker and interested in language, I did not know the word "tsunami" until 26 December 2004. From then on, I have always pronounced the [t]. Before then, I pronounced it "tidal wave".

  32. D Sky Onosson said,

    March 11, 2011 @ 2:51 pm

    It's not just the initial t-. Surely soonamee would be a far more anglicized form.

  33. Stephen Nicholson said,

    March 11, 2011 @ 2:54 pm

    All morning on the radio I've been hearing them say "sunami". Up to and including "a 'sunami'.

    In order to find out how people really say it in casual conversation, you'd have to record them in conversations when they don't know that their being recorded, or at least not conscious of the reason for the recording. Otherwise, you get biased results.

  34. Josh Treleaven said,

    March 11, 2011 @ 2:55 pm

    Another comment: in my reading, the initial "t" hardens (lengthens? lowers?) the vowel. I have no idea if this is closer or farther from the Japanese pronunciation, but sunami reads SUN a me, while tsunami reads [t]soo NAEM ee, or [t]soo NOM ee.

  35. Anthony said,

    March 11, 2011 @ 3:00 pm

    John Lawler – I'd also test "The Japanese tsunami", "the worst tsunami", and "the 2011 tsunami", to test when tsunami is preceded by /t/, /s/, and another consonant. In my own speech, the /ts/ is still present in 'tsunami', but I am linguistically elite.

  36. Mntnr said,

    March 11, 2011 @ 3:02 pm

    This Texan-American grew up using the /ts/ pronunciation. My two favorite dictionaries AmHeritage and OxAm both give that pronunciation as well.

    For tsetse and tswana I use /ts/; as for tsar I pronounce it with it initial z and spell it czar.

    I grew up with the expectation that English spelling was rationally phonetical—which has lead to a strange idiolect. I wish other English words were as easy to sound out as tsunami.

  37. Ice Wolf said,

    March 11, 2011 @ 3:07 pm

    @John Lawler

    Simple test: speak, elicit, or record a phrase like "There's been a tsunami", making sure to say "a tsunami"; an initial vowel should make any [t] before the [s] stand out.

    To me, i realise "a tsunami" in that sentense as [a.su:]
    And "Tsunami's are knows an tidal wave" as [tsu]
    "The black tsunami" as [k.tsu]

    Utterance initial the intrusive t seems to be more percievable at utterance beginning, and after stops. In other cases, I seem to be realising it not as my regular /s/ as in soon, or soup, but — here's where I lack the ability to explain — somewhere more palatal, like tongue in the 't' position, but saying s, it's more flat and closer to a /ʃ/ but still clearly an allophone of /s/.

    Though after reading contrived texts talking about "tsunami" I am now changing my thoguht to tsunami has an intrusive t that is more noticable after a pause/stop.
    I do speak Japanese well now, and the sound I use in English isn't the same I use dot realize つ. For that, it's it's more of a /t/ sound that has a forced 's' from the pronunciation of /ɯ/

  38. dw said,

    March 11, 2011 @ 3:07 pm

    John Wells's Longman Pronunciation Dictionary (first ed., 1990) gives /tsu'nɑːmi/ as first choice for both British and American pronuncation. He gives initial /s/ as an alternative.

    He also gives initial /ts/ as preferred for tsotsi, tsutsugamushi and Tswana, but not for tsar/czar.

    I guess he must be "elite" as well :)

    I agree with you about the general insaneness of English orthography, but I think you've picked the wrong target in this case.

  39. Leonardo Boiko said,

    March 11, 2011 @ 3:11 pm

    @Josh: the Portuguese word is something like “seaquake” (terremoto → maremoto), for both the quake itself and the waves. Though currently “tsunami” seems more popular in Brazil (with a clear initial [ts] even though it’s not a Portuguese cluster, initial or otherwise).

  40. dw said,

    March 11, 2011 @ 3:22 pm

    In English the word is pronounced [sunami] rather than [tsunami] since English does not allow syllable-initial [ts].

    English does not generally allow syllable-initial /ʃl/ or /ʃm/, yet they are readily found in Yiddish loanwords such as "schlepp" or "schmuck". As far as I can tell, these pronunciations are not confined to elites.

  41. Sid Smith said,

    March 11, 2011 @ 3:23 pm

    I got an eggcorn.

    My wife is Japanese and we were swapping emails about our friend in Yokohama. She wrote: "Don't worry. Yokohama is not the epic centre."

  42. Ralph Hickok said,

    March 11, 2011 @ 3:26 pm

    Maybe my experience is atypical, but in both my native Wisconsin and now in Massachusetts I've known families of Greek origin whose surnames begin with "Ts." They pronounce the initial "T" and so do their acquaintances. Going from the proper pronunciation of, e.g., :"Tsouprake" to "tsunami" is no big leap.

  43. Rube said,

    March 11, 2011 @ 3:38 pm

    Wow, I had no idea anybody pronounced the "t". Middle-aged, Canadian, move in educated circles, watch a lot of documentary style television.

    I am fascinated the extent to which,in this allegedly globablized time, people's perceptions of "normal" can still differ.

  44. michael farris said,

    March 11, 2011 @ 3:41 pm

    I usually say sunami, with initial [s] even though with Polish as a second language(and some knowledge of German and Hungarian) initial [ts] is easy enough.

    What I strongly dislike is that the Polish media is using tsunami instead of respelling it as cunami (Polish spelling makes much more sense, why muss it up with etymological spellings?). I also wouldn't be surprised if many Polish speakers think of it as an English (and not Japanese loan).

    Ironically, in Japanese it's not /tsunami/ but /tunami/ since the affricate is just an allophone of /t/ conditioned by the following /u/.

  45. Aaron Toivo said,

    March 11, 2011 @ 3:45 pm

    Mr. Poser: I submit that in cases such as these we add to the difficulty of learning to write English, and of remembering how to do so, for a negligible advantage.

    I submit that such difficulty is probably rare among native speakers of English, because tsunami is not the sort of word we employ on a daily basis and is probably much more likely to be first encountered in print than in speech. I have never heard of anyone asking how to spell /sunami/. No, what they always ask is how to pronounce . So it's not a word we normally have to learn to spell in an odd manner, but rather a word we already know how to spell but which has a silent letter in front, which we already deal with regularly in Greek loans (psycho) and even native words (knight) and is generally a non-issue.

  46. Chandra said,

    March 11, 2011 @ 4:12 pm

    @Aaron Toivo – "generally a non-issue":

    Tell that to the classes of often highly-intelligent dyslexic adults I teach, who want to be able to write at a level which reflects their vocabulary knowledge and speaking abilities, yet who encounter such seemingly arbitrary rule-breakers at nearly every second word.

    It's a common assumption that developed countries are highly literate and that most people can easily read a newspaper or write a letter, but the literacy figures suggested by research are quite shocking. I'm more familiar with Canada's statistics than with those of the U.S. or England, but I'd bet they're much the same. So please, from your vantage point as a well-educated, linguistically-capable individual, don't make the assumption that these kinds of things are "non-issues".

  47. Mntnr said,

    March 11, 2011 @ 4:21 pm


    I take Poser's argument to be that the English word is pronounced without a leading t. If, however, as there is evidence for, it is actually pronounced the way it is spelled—with an initial t—then it is not an exception.

    Do you teach the pronunciation with or without the t?

  48. Jim Breen said,

    March 11, 2011 @ 4:25 pm

    The New Shorter Oxford gives the pronunciation of the first syllable as "tsu", as it does with all 14 entries (mostly from Japanese) that start that way. Perhaps dropping the "t" is an American thing?

    It's interesting to read in that Wikipedia article that the first recorded use of 津波 was in a 16C chronicle.

  49. Chandra said,

    March 11, 2011 @ 4:30 pm

    @Mntnr: I wasn't necessarily advocating for or against any spelling of any particular word. To be honest, while reading this post I had to ask myself whether I pronounce it /tsunami/ or /sunami/, and I'm still not entirely sure. I'll have to try to catch myself saying it in context sometime.

    My point was merely that it's important not to dismiss concerns about the very challenging English spelling system (with its myriad exceptions, silent letters, and downright bizarre eccentricities) as unreasonable or unfounded.

  50. Chandra said,

    March 11, 2011 @ 4:36 pm

    …And for that matter, as someone else pointed out above, regardless of whether you pronounce the initial /t/, the spelling "tsunami" is still very much irreguar according to English spelling rules. Going strictly by the rules I've taught them, my class would probably try to render it as "tsoonammy" or something similar.

  51. Ellen said,

    March 11, 2011 @ 4:36 pm

    Frankly, I'm surprised how many people have posted how they pronounce it. Me, I don't think I've ever said the word. Not a word that comes up in conversation. Mostly, it's a word I've read. And probably occasionally heard on the news, but not often, as I usually only listen to local news, and live nowhere near a coast.

  52. Mntnr said,

    March 11, 2011 @ 4:51 pm


    Very good, I agree, and I hope I wasn't rude—especially how I have trouble with reading and pronunciation.

  53. marie-lucie said,

    March 11, 2011 @ 5:00 pm

    Just slightly off-topic: Bill Poser is undoubtedly familiar with the place-name Tsawwassen, the name of a seaside town in British Columbia where ferries depart and land on the way between Vancouver and Victoria. The initial "ts" of this Salishan name (stressed on the second syllable) is pronounced locally as [t], not [s] as one would expect from English speakers, let alone [ts] which the spelling suggests. (Query: does the "ts" represent yet another sound or sequence?)

    Johanne D: Gbagbo: the "gb" letter sequence does not represent a sequence of two separate consonant sounds (easy to pronounce in French) but a single sound (co-articulated g and b, pronounced at the same time), which is much more difficult to produce without training.

  54. Chris Waters said,

    March 11, 2011 @ 5:08 pm

    Hmm, I would tend to expect that omitting the "t" would be a prestige marker. In other words, my instinct is the exact opposite of Bill Poser's. I would expect a naive reading to produce the initial "t", and only someone with a more sophisticated knowledge of English convention would know to drop it. On the other hand, I can't entirely dismiss Bill's argument about bilingual pedantry (especially since I suffer from a bit of that myself), and it could be that pronouncing the "t" will end up being both a high and low marker.

    As for me, living on the US Pacific coast, I find both pronunciations unremarkable, and can't identify any personal preference. I seem to use either, interchangeably.

  55. Painni said,

    March 11, 2011 @ 5:09 pm

    I think 'sunami' would make the emphasis far less intuitive, though. 'Su-' isn't the most common initial set, either, and when it does appear it tends to be emphasized (super), if not pronounced with a different vowel entirely (sunny/suds/suffering). I'd argue that 'tsunami' earns its distinctive spelling even if the t isn't universally voiced.

  56. Chandra said,

    March 11, 2011 @ 5:17 pm

    @Mntnr: Not to worry, I didn't think you were being rude at all. Your point and your question were perfectly valid.

  57. Chris Waters said,

    March 11, 2011 @ 5:18 pm

    Hmm, both the online American Heritage (4th ed.) and Collier's Unabridged dictionaries list "ts" as the initial sound. Neither offers a plain "s" as an alternative. However, my older Webster's New Collegiate (c. 1980) lists the "t" as optional.

  58. Mark said,

    March 11, 2011 @ 5:21 pm

    Isn't this really the same sort of garbage word-rage rant that this site tends to come down on? I checked twice to see if this was an April Fools post that dropped 20 days early.

    How is this different than any of thousands of words we borrowed from the French where we kept their spelling or the fun variety of ways that we spell things like cerberus/kerberos as we thrash around for the best way to spell old Greek names? *ahem* "Chthonic" comes to mind, too.

    We borrowed a word that was originally spelled in a scholarly way and stuck with the weird spelling. Big Whoop.

  59. mollymooly said,

    March 11, 2011 @ 5:24 pm

    For me [tsunami] is the citation form and [sunami] the conversational form, just as I usually pronounce "asked" as [ast] rather than [askt].

    I would object to any proposal to respell "tsunami" as "sunami" just as I would object to any proposal to respell "asked" as "assed".

  60. Bloix said,

    March 11, 2011 @ 5:55 pm

    My dictionary has "tsoo-na-me" which is what I say, and I don't find the spelling odd at all. It's a conventional spelling for transliterated Japanese (not so different from sashimi, edamame, wasabi, etc.)

    And there's a big difference between sounds that aren't native to English but are easy to say, like "ts," sounds that a native English speaker can't pronounce without difficulty, like "gb," and sounds that simply don't exist in English and can't even be perceived without special training, like the sound signified by "f" in Fujiyama.

    It's odd that we don't have "ts" or "dz" in English, given that we have "ch" and "j." But perhaps because we do have "ch" and "j," "ts "and "dz" are not difficult for us to pronounce.

  61. J. Lee said,

    March 11, 2011 @ 5:58 pm

    FWIW Arabic seems to have borrowed the English spelling, with a /t/ — it's certainly not a native cluster, but I did hear an Arab newscaster pronounce the /t/. Although this affricate is not hard for English-speakers even initially (like "let's" in very rapid speech, or a phrase like "[I]t's all gone"), I confess that I have literally never heard it used in this word. Likewise I fail to see how someone could be unsure of their pronunciation of what would entail consecutive manners of articulation, i.e. some conscious effort if you speak English natively.

    I rather agree with Aaron Toivo that it is a non-issue; its odd form and low-frequency makes it much easier to remember, apart from marking it as an obvious borrowing, which can retain native spelling with impunity anyway. It's certainly not a good example of why English spelling is difficult.

    What Chandra should really be focusing on is not intelligent adults who are confused by exceptions, but university graduates who haven't written or read enough to internalize the use of homophonous there, their, and they're.

  62. Not a naive speaker said,

    March 11, 2011 @ 6:03 pm

    elite phenomenon: those who produce [ts] are a small minority familiar with other languages and of cosmopolitan orientation.

    Thank you Bill. Now I'm a member of a elite minority cosmopolitan group.

  63. Chris Waters said,

    March 11, 2011 @ 6:18 pm

    @Mark: I think this is different from the sort of prescriptivist poppycock this site regularly disparages. Bill isn't complaining because people aren't following "the rules"—he's complaining about the rules!

    That said, I think the various dictionary citations posted in this thread, all of which either allow or (in a couple of cases) require the "t" to be pronounced, shows that Bill was probably wrong about what the rules actually are, which is typical of prescriptivist nonsense rants. But since he was complaining about those non-rules, I expect he's happy to find he's wrong—something no rabid prescriptivist threatening violence on those who break his non-rules is ever happy to hear.

  64. M said,

    March 11, 2011 @ 6:20 pm

    It's odd that we don't have "ts" or "dz" in English, given that we have "ch" and "j." But perhaps because we do have "ch" and "j," "ts "and "dz" are not difficult for us to pronounce.

    I guess you left out "initial" here, since we have "cats" and "ads".

  65. Josh Treleaven said,

    March 11, 2011 @ 6:22 pm

    @Leonardo Boiko the Portuguese term is definitely more interesting to me, and more logical than what we've got in English, but you can see it still fails to distinguish between an earthquake under the ocean, and the resulting wave. I trust you and the rest of the LLers will help me spread my English coinage (okay, I admit I'm probably not the first), and coin similar ones for all other languages. "Quakewave" ftw

  66. carla said,

    March 11, 2011 @ 6:30 pm

    Bill Poser wrote: "Sure, some English speakers pronounce initial [ts]. It is conceivable that this will spread to the point that initial [ts] will become a natural part of the English sound system. However, I strongly suspect that this is an elite phenomenon: those who produce [ts] are a small minority familiar with other languages and of cosmopolitan orientation."

    What struck me as overbroad in your initial statement was the unequivocal claim that "in English the word is pronounced [sunami]." It still does; even if those of us who have [tsunami] in our English are a cosmopolitan minority with some knowledge of other languages, it is still a part of our English, mutually intelligible with people who [sunami] in their English.

    I also don't fully understand the verb phrase "become a natural part of the English sound system". What does this mean? Does a foreign sound necessarily become a natural part of the English sound system by virtue of appearing in some loanwords? Has initial [shm] become a natural part of the English sound system because of its appearance in a number of loanwords from Yiddish (shmuck, shmear)? Is it not possible for [tsunami] to exist in some native speakers' English without initial [ts] becoming a natural part of the English sound system?

    These are genuine questions, not meant Socratically or argumentatively.

    It is true, of course, that I and the many others who self-report [tsunami] are also self-selected to be people with some knowledge of languages (since we are here reading language log).

  67. carla said,

    March 11, 2011 @ 6:30 pm

    Please disregard the last paragraph of the above; it is cut-and-paste drool that should have been deleted before submission.

  68. Mark Mandel said,

    March 11, 2011 @ 6:35 pm

    I mentioned this post and thread to my son, including carla's comment and my reply (both of us grew up in NYC and pronounce the "ts"). His immediate reaction: "Growing up in New York, you don't need this kind of tsuris [Yiddish for 'trouble']."

    Right on target! New York English, even of non-Jews, contains a noticeable amount of Yiddish, in which initial [ts] is common.

  69. Josh Treleaven said,

    March 11, 2011 @ 6:43 pm

    @marie-lucie re: Tsawwassen. I grew up in Victoria, and what I remember is about equal distribution of people saying Tawwassen and Sawwassen. Maybe it was just my ear, but I can't ever remember anyone pronouncing both sounds as ts.

  70. Julie said,

    March 11, 2011 @ 6:48 pm

    I remember the Great Alaskan Quake and tsunami of 1964, although I was very young. As a small child in a California fishing family, I watched the excitement: midnight phone calls, followed by a rush to secure the boat against the oncoming disaster.

    I am pretty sure that I learned both words, "tidal wave" and "tsunami" at that time, the former being the term I heard at home, the latter on television. I've always thought the beginning "t" of "tsunami" was pronounced, although now I wonder whether I would have noticed that not everyone did so.

    Today, for the first time, I think I understand why they were called tidal waves: not for any belief that they were caused by tides, but because they behave something like tides, coming in wave after wave.

  71. John said,

    March 11, 2011 @ 6:56 pm

    What amuses me is the widespread use of the phrase "tsunami wave" – i.e. "harbour wave wave". I've never been sure why tsunami replaced "tidal wave", I think it was affectation – tsunami sounds moe exotic and educated than tidal wave. It's true that the pehnomenon in question has nothing to do with tides, but then it has nothing to do with harbours either.

    In one piece today I saw "tsunami", "tsunami wave" and "tidal surge". I can't wait for "tidal tsunami wave".

    Oh, fwiw, I say /ts/, and I did even before I spoke Japanese.

  72. Lisa said,

    March 11, 2011 @ 7:00 pm

    Somewhat related, I saw this comment on a website about proposed GOP budget cuts for the Pacific tsunami warning center:
    "Great idea GOP tsunamis are rare so much so I do not think there is a word for them in Japanese."

  73. David Eddyshaw said,

    March 11, 2011 @ 7:20 pm

    Let us start spelling (and pronouncing) "very" as "fery". No more Norman elitism for me! Don't see any reason to try to make the "v" sound in such an unnatural position as word initial. What am I, French?

  74. Aaron Toivo said,

    March 11, 2011 @ 7:28 pm

    @Chandra: So please, from your vantage point as a well-educated, linguistically-capable individual, don't make the assumption that these kinds of things are "non-issues".

    Sorry to split hairs here a bit, but I did say "generally" a non-issue. I was quite aware that exceptions existed, and so I used hedging words in my post such as "probably", "normally", and "generally", with the intent to signal that I was not claiming to describe everyone everywhere. I still feel that what I said is generally – as opposed to universally – true.

    As for literacy skills, I sympathize with those who struggle with the oddball spellings, but English spelling is no worse than average for a European language, and native Anglophones may tend to have exaggerated ideas of how bad our spelling system is (for instance, 85% of common words conform perfectly to the usual expected rules). And it shouldn't be hard to formulate a rule for when not to pronounce initial stops (P T CH C/K), even just "when they appear before any other consonant than R L W Y" would be a good start. Whereas learning to write is a rote memory task for most words in any case, and the challenges here are a) much greater than just the oddball words and b) mostly unaddressable in a spelling reform without sacrificing cross-dialect readability.

  75. Eric Christopherson said,

    March 11, 2011 @ 7:38 pm

    /ʃm/ seems to be spreading in some speakers from Yiddish-derived words to other English words. I don't recall ever encountering it in common words like _small_ or _smart_. Some examples I think I remember: _smear_, although that's complicated by the fact that there's a Yiddish borrowing _schmear_ (although I've never heard anyone use the "bagel spread" meaning that that entails). _Smarmy_ might be another. Certainly I've heard it in _smorgasbord_.

  76. Alex G. said,

    March 11, 2011 @ 8:05 pm

    Can't comment on this "insane" spelling – I've got neumonia and I'm late for my sychology class. No, wait, those are insane too. newmoaneeya. saikahlowgee?

  77. Qov said,

    March 11, 2011 @ 8:27 pm

    Josh, when I was a kid, people said Tawwassen, and I still do, but now it seems that most people say Sawwassen or [ts]awwassen. I've no idea how it's supposed to be pronounced in Halkomelem.

    I *think* I say [ts]unami but it's possible that I'm saying sunami and just thinking about the letter t as I start the word.

  78. J. Lee said,

    March 11, 2011 @ 9:08 pm

    yeah why all the english orthography bashing? I daresay it is easier to learn to write tsunami than 津波

  79. Bobbie said,

    March 11, 2011 @ 9:25 pm

    I had never heard the word pronounced (or knew its meaning) until I visited Hawaii in 1972 and saw the instructions in the telephone book on what to do if there was a possible tsunami. So I guess I always thought the word should be pronounced with an initial TS.
    (How quaint it sounds to talk about emergency instructions in a **phone book! )

  80. Thomas Westgard said,

    March 11, 2011 @ 9:29 pm

    J. Lee: hilarious! I agree.

  81. marie-lucie said,

    March 11, 2011 @ 10:32 pm

    lisa I saw this comment on a website about proposed GOP budget cuts for the Pacific tsunami warning center:
    "Great idea GOP tsunamis are rare so much so I do not think there is a word for them in Japanese."


  82. Faith said,

    March 11, 2011 @ 10:34 pm

    @ marie lucie — people trying to be fancy (CBC announcers, etc.) pronounce it "Sawasen" but regular folks tend to say "Tawasen". I have no idea which is correct, but one is more elevated.

  83. Chandra said,

    March 11, 2011 @ 11:11 pm

    @Aaron Toivo: If you will have a look at the statistics I posted a link to, you'll see that 48% of the population falls below the minimum literacy level considered necessary to deal with simple, daily reading and writing tasks, and a further 33% only meet that bare minimum. This gives us 81% of the population who struggle significantly with reading and writing. I hardly think such a figure justifies your assertion that the challenges of English spelling are "generally a non-issue". When you talk about words that "we already know how to spell", you are assuming the common misconception that the "we" you are referring to encompasses the majority of the population, when in fact it refers to a relatively small, elite percentage.

    Furthermore, it is not simply a matter of a few "oddball words". English spelling is well-known to be significantly more idiosyncratic, and contain significantly more exceptions, than most if not all Indo-European languages. Though it may be the case that 85% of English words follow the rules (and I question the accuracy and source of that figure), the fact is that the rules themselves are numerous, complex, and confusing. Asserting that it shouldn't be hard to formulate yet another rule indicates to me that you have very little understanding of the very real, very daunting struggle that so many people face when trying to learn to write in this language.

    I am not advocating spelling reform, mainly because I think such an effort would be futile. But it would be nice to see wider recognition in our society of just how difficult a struggle this is for such a large percentage of the population.

  84. fog said,

    March 11, 2011 @ 11:37 pm

    I would swear that I NEVER heard anyone pronouncing a "t" at the beginning of tsunami until today! I was 14 years old when the tsunami happened the Indian Ocean, so I was definitely old enough to already know the word and then hear it many times as I watched the news. Today a professor pronounced the "t" during a class discussion, though, and it sounded really strange to me. Now I come here and read all the comments saying that the "t" is completely normal for many English speakers. How strange.

    (And of course, how tragic the actual event is.)

  85. Brett said,

    March 11, 2011 @ 11:59 pm

    I pronounce it "tidal wave," always, for two reasons. The first is that "tidal wave" is the native English term for the phenomenon, and I have a certain fondness for it on that basis. The second reason is that the usual objection to it as "inaccurate" is baseless.

    I remember being pedantically instructed by my third grade teacher that we should use the word "tsunami" instead of "tidal wave" (and reading the same thing in a science magazine for kids a year or so later). However, I intuitively recognized this as the etymological fallacy even when I was nine, and I refused to give up "tidal wave." I used both terms interchangeably for a while, until I saw a crude video of a tidal wave's approach when I was a teenager. It's quite remarkable to see, the way the tide goes way, way out, before the wall of water smashes against the shore.

    A "tidal wave" does have nothing to do with the lunar (or solar) tide. But that's not the only meaning of the word "tide"; it can refer to other risings and lowerings of water levels, reminiscent of gravitational tides. I've even come across it used to refer to gradual changes in water level in inland bodies, for reasons that have nothing to do with the moon. Such usages are certainly less common, but that's hardly a surprise. The lunar tide goes in and out twice a day along every seacoast on Earth; other similar-looking movements are less common, and so they are less discussed. But they are, linguistically speaking, "tides" nonetheless.

  86. hector said,

    March 12, 2011 @ 12:45 am

    I've always pronounced the "t" in "tsunami." Aside from the fact that's the way it's spelt, I like the "ts" sound. It's fun to say. And in my Canadian English dialect, it's easy to say, so why not? (Maybe some English dialects have trouble pronouncing the "ts," I don't know).

    "English does not allow syllable-initial [ts]" strikes me, a non-linguist, as a bizarre statement. English, then, is a demi-god, that has the power to allow or disallow sounds used by us mere mortals? I'm always taken aback, and a bit fearful, when academics impute agency to the abstractions they study.

    "The word could perfectly well have been borrowed into English as sunami." It could have been, but it wasn't. Given that it wasn't, what harm has been done? "You say sunami, I say tsunami" — I mean, really, where's the problem?

  87. Alex G. said,

    March 12, 2011 @ 1:00 am

    @Brett – Since you admit that you made up your mind about this when you were 9 it's clearly not productive to argue with you, but surely one important thing about the word "tide" is that it originally meant *time*, and thus deeply implies regularity and predictability.

    Which in situations like today's makes "tidal wave" a term that has far less clarity, impact and gravity than "tsunami". Right?

    It'd be like insisting on calling eclipses "surprise sunsets". Why not use a term that implies an extraordinary disaster instead of an everyday regularity?

  88. John said,

    March 12, 2011 @ 1:07 am

    > Why not use a term that implies an extraordinary disaster

    Excellent idea. So where does "tsunami" come into it?

    As for appealing to ur-entomology… "nice" has roots meaning "excessively fussy", while "fornicate" has a root meaning "oven". Not sure where this gets anyone…

  89. John said,

    March 12, 2011 @ 1:08 am

    Of course that should be "ur-etymology", though ur-entomolgy could explain the bugs in my typing. It's late here.

  90. m.m. said,

    March 12, 2011 @ 1:14 am

    I completely forgot about "tidal wave", which is/was? fairly common here in coastal california, par the surfers.

    It's always [s] for me, and I can't recall it being said to any noticeable degree with [ts], even by the japanese americans I know. Hearing it with [ts] sounds more 'japanese' or 'culturally educated thus proper' english.

    Uhg, english orthography. Burn it with fire and nuke it from orbit. It's hard enough if you're a fully low back vowel merged speaker, but throw in a vowel shift that doesn't affect all vowels equally, and other regionalisms, forget it. Back to pre-standardization for me xD

  91. Alex G. said,

    March 12, 2011 @ 1:42 am

    @John – I'm not suggesting everyone keep in mind the ancient fact that "tide" meant "time" at all points – I guess I was trying to show off or something, should have just said that the word "tidal" always has had, and I think today still has, a strong connotation of regularity. It does, doesn't it? So all I was trying to say was:

    1) In English, "tidal" almost always refers to something very predictable and cyclical

    2) On the other hand, I think most English speakers would agree that the imported, exotic Japanese term "tsunami" refers to a non-predictable, non-cyclical, extra-ordinary event.

    3) Therefore, when describing a one-time huge crazy monster wave, I'd personally recommend discarding the term that is boring and confusing, and going for the more clear, exotic and vivid one.

    Sorry if the way I tried to say that was confusing, and I'd be interested to know if you disagree with those 3 points.

  92. pfisk said,

    March 12, 2011 @ 1:48 am

    I'm with Brett. Nothing wrong with "tidal wave." It's a wave that is a tide in and of itself. The argument for "tsunami" as "more accurate" is a silly.

    To the original point, I suspect Bill wishes he had checked for data before asserting his pronunciation premise.

  93. Pekka said,

    March 12, 2011 @ 4:14 am

    I think you will find a language or two that use the spelling sunami from the Wikipedia page's sidebar links. If you are willing to trust the Wikipedias of various languages, you find some interesting variations of the word.

    Some orthographics have the letter c represent the appropriate affricate sound, so they spell it cunami.

    In some languages they use a different spelling in place of the u, and in some language where they could do this they do not. For example, Afrikaans has tsoenami, but Dutch has tsunami.

    In Icelandic they have Flóðbylgja from native roots, which is what everyone should obviously adopt, whether they speak Icelandic or not!

    Wikipedia: Tsunami

  94. Pflaumbaum said,

    March 12, 2011 @ 4:17 am

    There's a tasteless joke been doing the rounds for the last few years, a pun on tsunami and Toon Army, which is what fans of Newcastle United Football Club call themselves (toon = town in Geordie dialect).

    The pun relies on [tsuːnɑːmi] sounding like [tuːn.ɑːmi], and becomes very weak if it's pronounced [sunɑːmi]. So for a lot of English English speakers I'm assuming the /ts/ is there.

  95. Colin Reid said,

    March 12, 2011 @ 5:02 am

    Seems to me that for the last few centuries, when English has borrowed words, it's tried to keep the original spelling (using some transliteration convention if necessary) and then English speakers give an approximate pronunciation using sounds they are comfortable with, and possibly the spelling takes them away from the original pronunciation. (What proportion of English speakers find word-initial 'ts' difficult to say?) By contrast, in Japanese, first an approximation to the sound is made in the Japanese pronunciation system (often simplifying consonant clusters if they make the word too difficult for a Japanese speaker to pronounce) and then the word is written phonetically. I don't see why either convention is inherently more 'sane'.

  96. Brad said,

    March 12, 2011 @ 8:30 am

    I have to agree with Colin Reid's suggestion, but I'd like to add something else, too.

    I don't know how it was for anyone else, but I had a terrible time relearning a distinction between the 'tsu' and 'su' sounds in words. I'm quite confident that if I had heard someone pronounce 'tsunami' with the 'tsu' blend when I was a teenager, I would have only perceived it as a strange way of pronouncing the initial s.

  97. army1987 said,

    March 12, 2011 @ 9:15 am

    Etymological discussions on which term is “better” seem silly to me: those waves don't have any more or any less to do with harbours than with tides, AFAICT, and anyway according to the etymologies awful and awesome ought to be synonyms whereas they are antonyms.

    In Italian tsunami is pronounced /tsu'nami/, but /ts/ is a perfectly normal syllable onset in Italian. They used to be called maremoti (“seaquakes”) but for some reason they aren't anymore. I don't think I've ever pronounced this word in English, but probably I'd go with /su'nami/ (where u is the last vowel in INTO, a is the vowel in TRAP (which is nearly identical to the one in PALM in my accent, anyway) and i is the last vowel in HAPPY).

    I also don't fully understand the verb phrase "become a natural part of the English sound system". What does this mean? Does a foreign sound necessarily become a natural part of the English sound system by virtue of appearing in some loanwords? Has initial [shm] become a natural part of the English sound system because of its appearance in a number of loanwords from Yiddish (shmuck, shmear)? Is it not possible for [tsunami] to exist in some native speakers' English without initial [ts] becoming a natural part of the English sound system?
    Many languages have much more lenient phonotactics for loanwords than for native words, but English doesn't appear to be one of them: witness what happened to the initial consonant of words such as psychology. There once was an argument on a WIkipedia talk page that /sf/ and /skl/ shouldn't be considered valid English syllable onsets because they only occur in words of Greek origin, but I pointed out that the fact that sphere unlike psychology, pneumonia, etc. managed to keep its initial consonant cluster even if it's otherwise fully assimilated means that there's nothing in the English phonotactics forbidding initial /sf/. (Also, all syllable onsets of the form /s 0/ + /p b t d k g/ + /l r 0/ in which /s/ doesn't precede a voiced stop and /l/ doesn't follow an alveolar stop are allowed, saying that /skl/ is not just because there don't happen to be any words of Germanic origin with it seems to violate Occam's razor to me.)

  98. Keith Ivey said,

    March 12, 2011 @ 9:33 am

    Isn't initial [ts] perfectly normal in conversation in many English dialects? It certainly seems to occur in my normal casual pronunciation of sentences that start with "What's" or "Let's". Isn't "Tsup?" for "What's up?" widespread?

    For that matter, it can even be followed by another consonant, as in "Tsko" for "Let's go".

  99. Ellen K. said,

    March 12, 2011 @ 9:44 am

    @Chandra: Aaron Toivo was talking specifically about the word "tsumani", and saying that that specific word is "generally a non-issue" with regards to the difficulty of learning to write English.

    So, even if what you say is valid, it's not a valid argument against what he says.

  100. language hat said,

    March 12, 2011 @ 10:15 am

    I am surprised that you say it is pronounced [sunami] in English – I always pronounce the initial t, and in my experience most English speakers do as well.

    Same here. I agree with Mark: this is a very odd word rant to find at the Log.

  101. Ellen K. said,

    March 12, 2011 @ 10:18 am

    Now after listening to a sound sample with the [ts] (in this morning's post from Mark Liberman), my thinking is, if I learned the word (or another word with ts, used in English) by listening to it, I would probably internalize it as beginning with a plan /s/, simply because /ts/ is not part of my listening sound inventory. It's easy to see the word "tsunami" and think of it as having a t-sound. Harder to hear and and hear it as starting with /ts/.

  102. Ellen K. said,

    March 12, 2011 @ 10:23 am

    I'm surprised that two people posting here claim to see the initial post as a rant. I don't see that. There's no evidence in the post for taking it that way. "Insane English spelling practices" is admittedly unscholarly, but frankly expressing an opinion on English language spelling does not make the post a rant.

  103. Jerry Friedman said,

    March 12, 2011 @ 11:15 am

    @Eric Christopherson: I suspect the pronunciation of "smorgasbord" with /ʃ/ is from true or false ideas of how it's pronounced in Swedish.

    The only word I can think of where an initial /ʃ/ seems to have "spread" is swag, sometimes pronounced with /ʃw/ when it means stuff given away by exhibitors at trade shows, employers at company picnics, etc. A search for shwag and schwag will turn up examples.

  104. John said,

    March 12, 2011 @ 12:32 pm

    Czar/tsar is not a good parallel because it was routinely spelled the second way. I seem to remember a deliberate shift in the US during the 1970s, but I may be imagining that. The OED lists several 19th c. British examples of tsar.

  105. John said,

    March 12, 2011 @ 12:32 pm

    Sorry, that should be the "first" way!

  106. HP said,

    March 12, 2011 @ 3:17 pm

    @Pekka: I did the same trick with the Wikipedia "Languages" sidebar. What I found interesting is that, at least for the links I clicked through, the article always lists the word they used to use in that language before adopting the Japanese word. E.g, vloedgolf, maremoto, etc.

    Oddly enough, although the Japanese word isn't inherently more correct than any other term, it seems to have displaced the local word almost universally.

  107. Barry said,

    March 12, 2011 @ 3:27 pm

    Sue Nami is a pretty minor faux pas compared to Carry Okie.

  108. Chandra said,

    March 12, 2011 @ 3:34 pm

    @Ellen K: I disagree. Mr. Toivo's original post was directly challenging Bill Poser's statement that "in cases such as these we add to the difficulty of learning to write English, and of remembering how to do so, for a negligible advantage." His reply was that "such difficulty is probably rare among native speakers of English", and went on to explain why he thought so, giving tsunami as an example case. I was defending Mr. Poser's point of view.

  109. maidhc said,

    March 12, 2011 @ 6:56 pm

    In Lancashire dialect, phrases like Monty Python's "trooble doon at t' mill" can be heard by some people as lacking an article. But to me it sounds as though "the" has been reduced to "t" and then tacked on to the front of the following noun. So phrases like "at t' pub" would be pronounced with an initial "tp", "'t mill" "tm" and so on, although it's written with "'t" separate.

    And surely you can't get much more English than Lancashire.

    In the US, there was a potential presidential candidate Paul Tsongas, whose name did not create any big controversy.

  110. Rubrick said,

    March 12, 2011 @ 7:29 pm

    While I would stop short of describing this post as a "rant", it does seem awfully short on fact and long on peeve for a LL post. "In English the word is pronounced [sunami] rather than [tsunami] since English does not allow syllable-initial [ts]" — Mark Liberman surely wouldn't have just baldly stated this as fact; he would have run a Breakfast Experiment™ and checked. The results might have been interesting.

    Bill Poser's defense, when confronted with readers who at least believe they pronounce the [ts] — "I strongly suspect that this is an elite phenomenon: those who produce [ts] are a small minority familiar with other languages and of cosmopolitan orientation" — seems bizarre coming from a linguist. Since when is "elite English" no longer a legitimate variety of English? Should only the pronunciations of, say, working-class American Midwesterners be considered a part of the language? If a certain pronunciation prevails among one socioeconomic (or regional or educational) group, that's something worthy of study, not dismissal as irrelevant.

    Bill's characterization of English spelling practices as "insane" also seems unprofessional. Is he dismissing orthography as a field of study? Would he feel comfortable referring to "Luganda's insane gender system" or "!Kung's insane phonology"? English spelling may indeed be unsystematic, or nonconstructive, or an impediment to learning the language — these are claims that can be (and no doubt have been) tested. If English spelling is indeed an impediment, learning why and how it has managed to persist seems much more interesting than ragging on it.

    Bill's entitled to his opinions, of course, and even to post them online; but as a reader, I don't really need to go to a blog written by professional linguists to find posts of the "English spelling is stupid, and here's another example" variety.

  111. blahedo said,

    March 13, 2011 @ 7:21 pm

    I'm not going to weigh in on the s/ts argument, but I find some of the alternate proposed spellings to be interesting. As noted by at least one person, "sunami" reads differently (more like /'sʌ.nə.mi/, and while a few other proposals ("soonamee", "tsoonammy" etc) might elicit the right pronunciation, they struck me as weirdly archaic. Archaic? I probed that a bit and thought, well, not archaic generally, but Indian.

    I realised that what I was picking up on there was a shift in how English adopts foreign words. Once it would have been relatively common to borrow a word and come up with the spelling that best leverages existing English spelling practices to suggest the correct pronunciation. Now, though, it's much more common to either A) use the native spelling, if in the Latin alphabet, or B) use something approaching continental values of the various letters. So much so that seeing the older practice (leverage English spelling rules) comes across as patronising or mocking, as if to use dialect spelling. About the only place I run into such practices generally are at Indian restaurants*, so I'd be confused by a spelling like "soonamee" and I'd go looking for some sort of xenophobic or at least anticosmopolitan agenda.

    *I first learned the spelling of one common dish as "alu gobhi", but "aloo gobee" seems to be widespread too; it threw me the first time I saw it, almost as if the restaurant owner didn't know what they were doing—obviously not the case.

  112. Bob C said,

    March 13, 2011 @ 9:08 pm

    A tsunami can cause a lot of tsouris.

  113. Greg Morrow said,

    March 13, 2011 @ 9:23 pm

    I would have sworn I pronounced /ts/, but I tried "a tsunami" and "worst tsunami", and b'god, it's clearly an /s/ in ordinary pronunciation.

    But said in isolation, I'm pretty sure I have /ts/. That suggests that it's in the lexicon that way, and a production rule steps in and simplifies it when it's part of a stream.

  114. D Sky Onosson said,

    March 13, 2011 @ 10:19 pm

    @ blahedo: I proposed soonamee with tongue firmly in cheek. However I would note that a Korean friend of mine, whose given name was 수미 /sumi/, in fact wrote her name as Soomi even though standard Korean-to-English transliteration would have it as Sumi. I'm pretty sure she did this to make it easier for anglophones to get the correct pronunciation – though as a linguistics student at the time, I remember finding it really strange.

  115. John F said,

    March 14, 2011 @ 4:51 am

    I've heard a few shoonami's on Radio 4. Why not spell it with a z, which would be closer to tsu than su?

  116. Amy Stoller said,

    March 14, 2011 @ 8:23 am

    I don't pronounce the p in psyche. Sosumi.

  117. Mary Beckman said,

    March 14, 2011 @ 10:32 am

    I wonder if "shallows" or even "shoal water" may be a better translation of 津 in this case? cf. http://en.wikipedia.org/wiki/Wave_shoaling for a description of the phenomenon, and poems such as Man'yoshu 1188 for this possibly older meaning for the word.

  118. David Marjanović said,

    March 14, 2011 @ 4:23 pm

    Spelled Tsunami in German as well, not with z, the normal letter for [ts]. Hepburn romanization is treated like a foreign orthography, and recent loans strongly tend to keep their original spellings in German.

    (…Except exceptions. Tokio, Kioto. BTW, also Kenia. *word rage*)

    FWIW Arabic seems to have borrowed the English spelling, with a /t/ — it's certainly not a native cluster

    Depends on which kind of Arabic. Standard Arabic lacks initial clusters completely, but the farther west you go from Arabia, the more vowels drop out; in Algeria and Morocco, a wide variety of initial clusters including [ts] occurs frequently.

    "English does not allow syllable-initial [ts]" strikes me, a non-linguist, as a bizarre statement. English, then, is a demi-god, that has the power to allow or disallow sounds used by us mere mortals? I'm always taken aback, and a bit fearful, when academics impute agency to the abstractions they study.

    Let me ignore the last sentence and focus on the one before. Can you begin a word, or even just a syllable, with [ŋ]? That's ng as in singing (not as in finger, which has [ŋg]). Chances are you find it quite difficult, because English (like almost all languages in Europe) "doesn't allow" [ŋ] to occur in such positions. Rules like these are collectively called phonotactics; the questions in this discussion are what the rules of English phonotactics are and how they are changing.

  119. Mel Nicholson said,

    March 15, 2011 @ 7:57 am

    I did an experiment where I recorded yesterday's water cooler talk and avoided using the word myself. It was pretty easy to get sample just by asking "What do you think about Japan?"

    My impression at the time was that /ts/ versus /s/ was fairly even, but the recordings didn't back me up on that. Overall there were 15 people who used /s/ to only 6 who used /ts/. 5 of the 6 who used /ts/ were native speakers of Mandarin and the sixth was a native speaker of a Georgian language. None of the native english speakers used /s/.

    Next I want to look at the word "soon" and digitally splice out a minimal pair to see if there is anything in the vowel that would explain my earlier perception.

  120. Heather Smith said,

    January 21, 2014 @ 4:02 am

    The ENGLISH Oxford dictionary gives the pronunciation as ts. It would seem that at least some of you are USglish (US for United States). As for the critical French person, words like psychology and pterodactyl have Latin or Greek origins, hence the pronunciations. Every language has unique pronunciations for different groups of letters even yours.

RSS feed for comments on this post

Leave a Comment