Matrix in Japanglish: why, why, why?

« previous post | next post »

Lareina Li called my attention to a delightful clip from the Matrix trilogy as dubbed in Japanese accented English. But before you watch it, try listening to the sound track to "see" how much of it you understand without looking at the subtitles.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

If you missed any of it, you may want to try the subtitled version.

After viewing this video several times, the following somewhat random questions and observations come to mind:

1. How intelligible is this type of Japanized English to someone who is unfamiliar with the genre?

2a. Is this form of Japanglish entirely consistent with the phonological rules of Japanese?

2b. Or are there (slight or systematic) accommodations made to English phonology?

3. If 2a. is true or nearly true (as I suspect it to be), is this phenomenon of speaking English strictly within the phonology of Japanese peculiar to Japanglish, or does it happen with many other languages as well? I may be wrong about this, but it seems to me that the phonological conventions of Japanglish are more fixed and "standardized" than for similar versions of English elsewhere.

4. Would it be possible for Japanese to converse freely in Japanglish?

5. This reminds me of a French friend whom I've known for over three decades. When we are speaking together in English and he becomes excited and starts to speak quickly, his French accent is so thick that I can only understand about 30% of what he is saying, yet he is speaking English (French accented English, that is, which I consider to be quite different from Franglais or Frenglish).

To what extent do we have such more or less fixed forms of English accented according to the phonology of other languages (French, Russian, German, Turkish, and so forth)? Indian English seems to be a somewhat different case, since similar version of this accent seem to be shared by speakers with a variety of different native languages.

6. Finally, and most specifically, the insistent "why? why? why?" — pronounced HWAI (with the aspiration at the beginning particularly prominent, especially in the fourth utterance) at the beginning of the clip is especially intriguing to me, since — to the best of my knowledge — this is not a syllabic form characteristic of Japanese per se, though it would be perfectly natural in Chinese. So how does HWAI get generated from English "why" in accordance with the phonology of Japanese?

The katakana transcription of "why" is ホワイ(ho wa i); "where", ホエア (ho e a) or ホエール (ho ē ru), and "what" is ホワット (ho wa t to). For some reason, the Japanese place the sound "wh-" into the ハヒフヘホ (ha hi hu/fu he ho) syllabic letter group. This reminds one of the hookah-smoking Caterpillar in Alice in Wonderland, who is very fond of "wh-" words (though being an old-fashioned British caterpillar, he presumably pronounced them as "hw-" words), and whose most memorable sentence is exhaled as "Whooo … are … you?". (Of course, who is not really relevant to this discussion, since the standard pronunciation simply begins with /h/, shared by everyone who lacks more general h-dropping.)

Here is how Hiroko Kimura Sherry remembers Disney's version of the Caterpillar:

The Caterpillar, while smoking a long hookah on the top of a huge mushroom, asks many WH-questions to Alice, such as "Who are you?", "What do you mean by that?", and "Why?" This is my most favorite episode in Alice in Wonderland, and when I saw this scene in Walt Disney's film, a lot of smoke came out of the Caterpillar's mouth (=fully aspirated WH-) as he spoke, and I still remember myself feeling that was a perfect way to pronounce those questions. It may not be natural, but it was a perfect interpretation of the scene. This is just a personal memory of mine.

Some Japanese dictionaries offer for "why" the alternative katakana reading ワイ (wa i), evidently following American custom (see below), although ホワイ(ho wa i) is much preferred.

In thinking about this problem of "why" becoming HWAI in Japanglish, I wanted to see how various English-language dictionaries describe the pronunciation of this word. The answer is that they are collectively rather confused.

Among American dictionaries, American Heritage gives the aspirated version first ([hwahy, wahy]) and offer a recording that is also aspirated;  Merriam-Webster gives the same order of symbolic pronunciations ( \'hwī, 'wī\), but offers a recording without aspiration; Encarta reverses the order of the alternatives ([ wī, hwī ]) (and its recording seems unfortunately to be missing).

Among British dictionaries, the OED gives only the aspirated pronuncation /hwaɪ/. In contrast, the Cambridge Advanced Learner's Dictionary gives only the pronunciation /waɪ/, and offers two recordings, one labelled "UK" and one labelled "US", both of which have initial [w] without aspiration.

I personally pronounce "why" as [waɪ], with an unaspirated [w], and thought that other Americans did so too.  But the aspirated wh- has a complex distribution in space, time, and culture.

The wikipedia article on the wine-whine merger offers this map (from The Atlas of North American English) showing "the area in which the contrast between the pronunciation of /hw/ and the pronunciation of /w/ is greatest", and observes that "In most other areas of the United States, the pronunciation has merged so that both sound the same":

However, there seems still to be something left of the prescriptive attitudes described in Raven McDavid et al.,"h Before Semivowels in the Eastern United States", Language 1952:

Although the pronunciation of /h-/ before vowels does not constitute a social shibboleth in the United States, there is evidence that the presence or absence of /h-/ in words like whip and humor is often considered a test of social acceptability. Thus when Thomas Pyles recently remarked that in his dialect (of Frederick, Maryland) the cluster /hw-/ does not occur, despite the efforts of well-meaning schoolteachers to impose it on generations of students, a reader immediately commented that nowhere had she observed a person of true culture who did not possess that cluster.

Pyles' sin was to observe, in an article in College English (1948) that

To this day "the baby whales" and the "the baby wails" sound exactly alike in my pronunciation, as they do in the pronunciation of many speaker in all parts of the country, cultured and uncultured alike.

McDavid et al. continue:

Such responses are not confined to laymen. T.R. Lounsbury and William Dwight Whitney […] have insisted that there is a social stigma attached to those who do not pronounce /h-/ in words of these types. H.L. Mencken, on the other hand, considers the pronunciation of /h-/ in whip etc. an affectation.

They quote Mencken (from The American Language, 1947) as follows:

The majority of Americans seem to have early abandoned all effort to sound the h in such words as when and where. It is still supposed to be sounded in England, and its absence is often denounced as an American barbarism, but as a matter of fact few Englishmen actually sound it, even in their most formal discourse.

A 1999 page on the Words@Random site observes that

Long considered the norm, "hw" was, until the mid-twentieth century, the only pronunciation shown in both American and British dictionaries. […]

It was not until the 1960s that virtually every major dictionary publisher in the U.S. acknowledged the increasing use of "w" in these words by showing both pronunciations. They are still doing so. However, many current British dictionaries, including the "New Shorter Oxford English Dictionary" (1993) and the "Collins English Dictionary" (1986), have dropped the "hw" entirely, showing only "w." For the most part, they have gone directly from "hw" to "w," eliminating the interim phase that shows both, the phase still characteristic of both American dictionaries and American speech. […]

As with so many of these fading pronunciations, the use of "hw" is fraught with sociolinguistic implications. People tend to have strong feelings about how things are said. Despite its increasing rarity, some people associate "hw" with educated, desirable, elegant speech. Those who use it are reluctant to let it go or see it go. And some who do not use it would like to emulate it. At the same time, as we have noted, it is in England, the home of the archetypally elegant Received Pronunciation, that "hw" is disappearing even more rapidly than in the United States.

Whatever its current distribution in geographical and social space, the  /hw/ pronunciation has clearly been adopted with a vengeance into  Japanglish.  And the aspiration in native English varieties that retain it seems to be slight (Disney's caterpillar aside) whereas it is very prominent in Japanglish.  Presumably this is because within Japanese phonotactics, [hw] is in fact somewhat unnatural, and must be rendered as a separate /hV/ syllable, with the V being /u/ or /o/ depending on context.

[A tip of the hat to Hiroko Kimura Sherry Nathan Hopson, and Miki Morita.]


  1. Jarek Weckwerth said,

    April 9, 2011 @ 6:01 pm

    The Cool Whip connection is obligatory, isn't it?

    [(myl) Indeed.]

  2. David Eddyshaw said,

    April 9, 2011 @ 6:20 pm

    I have voiceless w, rather than a [hw] cluster in my idiolect (due to an early childhood in Scotland, despite the fact that my sound system is mostly British RP. ) It's not unlike the sound written f in the Hepburn romanisation of Japanese. Maybe it is this sound which is at the back of the Japanese versions, rather than a consonant cluster? After all, sequences like [fa] in English get Japonised as fua- too.

  3. Russell said,

    April 9, 2011 @ 6:31 pm

    I would say that 2a is nearly true. If we limit "Japanese phonology" to what occurs in native and Sino-Japanese vocabulary, then there were some semi-consistent adaptations to English, including

    – lack of affrication of /t/ and /d/ before /u/ in some cases (but not all)
    – English /v/ as /v/ instead of /b/
    – what sounded like a rounded vowel for the high back vowel (or, it's just my California front vowels messing up my perception)

    The first two are potentially violated in non-Chinese loanwords, and have conventional orthographic representation, but in my experience are sometimes articulated with native phonology.

    FYI, there is a word for dress shirt, ワイシャツ (wa i sha tsu), from "white shirt." It has never, AFAIK, been ho wa i sha tsu.

  4. Jim Breen said,

    April 9, 2011 @ 6:32 pm


    >> 2a. Is this form of Japanglish entirely consistent with the phonological rules of Japanese?

    Entirely. I think the person was reading from a script written in katakana-ized English.

    >>2b. Or are there (slight or systematic) accommodations made to English phonology?

    None that I could hear.

    >> 4. Is it be possible for Japanese to converse freely in Japanglish?

    I don't think so. It's not a language of discourse.

  5. Dw said,

    April 9, 2011 @ 6:43 pm

    For what it's worth, I've noticed that a lot of highly educated American-born Chinese-Americans pronounce /hw/.

  6. Carl said,

    April 9, 2011 @ 7:08 pm

    I'm not convinced waishatsu comes from "white shirt." I think it's a play on T-shirt, the Y-shirt. :-) White is pretty consistently transliterated as howaito. A quick Google shows 500,000 hits for waito (mostly about the Isle of Wight), and 170m for howaito.

  7. Carl said,

    April 9, 2011 @ 7:11 pm

    FWIW, Japanese Wikipedia buys into the White Shirt theory. Oh well.

  8. Peter said,

    April 9, 2011 @ 7:13 pm

    > 1. How intelligible is this type of Japanized English to someone who is unfamiliar with the genre?

    Reasonably so. I have no experience with this genre, and little familiarity with Japanese-accented English. I attempted to transcribe it before reading the subtitles — not too much effort, just a couple of passes. I ended up with two lacunae (I couldn’t get anything for “Is it freedom or truth?” or “Because I choose to!”) and two errors (“something good” for “something”; “forever” for “for love”).

    It was certainly more intelligible than plenty of strong US/UK regionally-accented dialogue in films — although, of course, such dialogue is usually not as hyper-articulated as this is.

  9. kuri said,

    April 9, 2011 @ 7:14 pm

    I think there are a few slight accommodations to English phonology. Listening very carefully, I think he did some things like sometimes dropping the terminal o sound in it/itto and changing the sh sound in insipid/inshipido to and s sound.

  10. Rob said,

    April 9, 2011 @ 7:31 pm

    Having lived in Japan for two years, and had a Japanese SO for 2 more, I'm probably a poor point of data, but:

    1.) I got almost everything that was said (though the faster it got the more likely I might have, out of context, mistaken it for actual Japanese. That said, my job was teaching conversational English to Japanese people, so I have practice with this.

    2.) It is, on the whole, consistent with Japanese phonology. One of the tell-tale features for me are

    a.) the u vowel appended to the ends of English words. (fai-ti-n-gu (fighting); sa-mu-thi-n-gu(something)) (normal Japanese phonology only allows as a final consonant,and foreign words are often converted with a (sometimes unvoiced) u.

    3.) It is, on the whole, VERY standardized, and heavily based on spelling (which may help explain the HoWAI.) (witness: the MATRIX (pronounced "mah-to-ri-ku-su" with an "ah" sounds in reflection of the English *spelling* though in variance with normal English pronunciation.) the pronunciation of words ending in "ing" in English is uniformly recognized and standardized as -ingu"

    4.) My students frequently spoke had English conversation in this heavy of an accent, though their ability to speak was uniformly much weaker than their ability to understand each other. I expect that particularly devoted students of English might, either as practice or as a semi-secret language have conversations like this. The average Japanese person would have a very difficult time having a significant conversation in this kind of English.

  11. John said,

    April 9, 2011 @ 7:33 pm

    >> To this day "the baby whales" and the "the baby wails" sound exactly alike in my pronunciation, as they do in the pronunciation of many speaker in all parts of the country, cultured and uncultured alike.

    I think this and similar comments misses what is actually happening, at least in my ideolect, and perhaps in my dialect group as well (the SE Colorado of 60 years ago):

    Yes, I usually drop the "h" when I say "the baby whales" unless I am speaking slowly and carefully. On the other hand I would never drop the "h" when I say "Whales are common." (that is, at the beginning of a sentence or clause), even when speaking fast and casually. For me there is a big difference between the sentence/clause initial context and the medial context. I have no data, but my hunch is that this contextual difference is important for many (but not all) other native English speakers as well.

  12. Yuji said,

    April 9, 2011 @ 7:45 pm

    As a Japanese educated in Japan, I can testify that, at least twenty years ago, I was taught that the word why is pronounced with [hw]. The dictionary I had had [hw] as the normative pronunciation. I was surprised when I came to the States a few years ago to find that they mostly pronounce "why" with just [w].

  13. Coby Lubliner said,

    April 9, 2011 @ 8:02 pm

    Matorikkusu probably comes from German, not English.

    Shatsu for 'shirt' (like katsu for 'cutlet') goes back to the Nagasaki era; in the local dialect, I believe, tu was (is?) pronounced [tu], not [tsu]. The modern Japanization of 'shirt' would be shaato.

  14. disfraz said,

    April 9, 2011 @ 8:34 pm

    Re: ワイシャツ (wai shatsu)–if this is indeed from "white shirt", it's interesting that the (unofficial) holiday White Day is ホワイトデー (howaito de:), with the aspiration well and truly present.

    And fwiw I'm another person with a reasonable amount of experience with Japanese and Japanglish who found this 90% intelligible on the first listen, and completely fine on the second. As for conversation, I think if one speaker had an accent like this and the other knew what to expect (was an EFL teacher, for instance) then there wouldn't be much of a problem. Not sure what would happen if both sides of the conversation had accents this strong.

  15. Don Sample said,

    April 9, 2011 @ 8:39 pm

    Is this an attempt to dub The Matrix, or is it video of a Japanese man practicing his English by reading along with the subtitles as he watches the movie?

  16. Ethan said,

    April 9, 2011 @ 9:22 pm

    I've never studied Japanese and very rarely (if ever?) heard heavily accented English spoken by a Japanese; I'd say there were about 8 spots that weren't clear to me on the first listen.

    As far as "wh": I'm from suburban Connecticut and I've only know one person, a middle-aged co-worker, who pronounced it as "hw." I always thought it sounded affected.

  17. Ethan said,

    April 9, 2011 @ 9:31 pm

    And how could I forget: The Cool Whip scene has nothing on this classic film scene about the hw- sound. (You can be forgiven for not having seen the entire film.)

    My safe word will be hwiskey.

  18. David Moser said,

    April 9, 2011 @ 9:39 pm

    I've noticed that English teachers here in China will often use the /hw/ pronunciation when teaching the Chinese kids what they consider the correct pronunciation of words like "why" — but then of course they revert to /w/ when speaking normally. The kids are confused and just end up ignoring the distinction. As to Japanglish pronunciation, a few of us in the past have tried to come up with some spelling conventions to render Chinglish pronunciation, but nothing ever came of it. The exercise was prompted by a puzzled Rachel DeWoskin ("Foreign Babes in Beijing") answering the question from a Chinese friend "Why you have so many fet peebo in America?" ("fet peebo" = "fat people"). I'm so engulfed in this kind of English these days to even notice it anymore.

  19. Rob said,

    April 9, 2011 @ 9:41 pm

    fixed: (normal Japanese phonology only allows n as a final consonant)

  20. Nanani said,

    April 9, 2011 @ 10:34 pm

    Commenting from Japan here.

    For question 1, I can't answer as I am far too familiar with this particular genre.

    I'd say 2a. (Yes to "Is this form of Japanglish entirely consistent with the phonological rules of Japanese?")

    I would bet money on the reader having a script in katakana (that is, English rendered into the Japanese syllabary, phonetically) and not actually reading English text.
    The giveaway is that words with non-transparent spelling, such as "could", are consistently pronounced without tripping on the silent letters.
    If he was reading English text and had the English proficiency to master silent letters, he would also have enough proficiency not to say "ing" as "ingu" with so much emphasis on the "gu".

    More succinctly, the deviations from English pronunciation are consistent with katakana renderings but not with semi-acquired English.

    3. I don't know.

    4. I highly doubt it. Certainly not outside an English conversation class. Actually communicating in this manner would require a grasp of the (very different) grammar rules, after all. Anyone who had a fluent grasp of English grammar wouldn't sound like THIS. (Still accented, probably, but not like this.)

    5. I leave it to the rest of the commenters.

  21. Russell said,

    April 9, 2011 @ 11:51 pm


    I wonder if the lack of h in "waishatsu" (temporarily assuming it has an origin in "white shirt," which I have no real evidence for; it's just what I was taught) is part of a phenomenon where some loan words came into the language in a way that more closely reflects how it was pronounced (or perceived) by whoever got the process started.

    For instance, there's also メリケン粉 (meriken-ko), despite the fact that elsewhere you have アメリカ (amerika). The fact that it's not ワイトシャツ or ワイツシャツ might point to a non-spelling-influenced loan process as well.

  22. Russell said,

    April 9, 2011 @ 11:56 pm

    Okay, just because it's interesting, apparently a way to refer to brass knuckles in Japanese is メリケンサック (meriken sakku), apparently after "American sack."

  23. Vasha said,

    April 10, 2011 @ 12:05 am

    I (American) had a second-grade teacher circa 1979 who tried to get us to pronounce the hw aspirated. Besides saying that this was the right way, she insisted that since the h was spelled it should be pronounced — an argument that sounds feeble even to a second-grader! Anyway, she had essentially zero success in changing anyone's pronunciation. Conversely, a visiting educator who ridiculed my pronunciation of aunt as [ænt] succeeded in embarrassing me enough that I said [ɑnt] for years, and it took a conscious effort to go back to using the same pronunciation as the rest of my family.

  24. Jerry Friedman said,

    April 10, 2011 @ 12:25 am

    [hw] is more common in northern New Mexico than anywhere else I've lived, i think, though still not common. A few older people with strong Spanish accents may pronounce it with friction like the beginning of Juan.

  25. Ron said,

    April 10, 2011 @ 1:13 am

    Somewhat related to question 4, if you've ever been in hospital in Hong Kong you will hear a very strange type of English being spoken by the doctors to each other (but not to other staff). A native English speaker who isn't familiar with the medical terms may not even recognize it as English through the heavy accents of some of them.
    The reason for this is that their training is all in English even though very few of them can speak it well. They need to revert to Cantonese to make jokes or any non-medical talk. So it's a sociolect and they aren't really conversing freely but perhaps this is one possible answer to that question.

  26. John Cowan said,

    April 10, 2011 @ 2:45 am

    My wife was born and raised in North Carolina, right in the purple zone above, and is quite unable to believe that the people around her (including her husband and daughter) don't consistently distinguish between /w/ and /hw/. I made a joke about "whining but not dining", and she heard it as simply "wining but not dining". Both of us can hear the distinction and can make it if we want to, but minimal pairs in actual use are few, and so my wife in ordinary conversation simply hears a distinction that does not in fact exist.

  27. michael farris said,

    April 10, 2011 @ 2:57 am

    I noticed something strange with initial wh- in my pronunciation. I tend to produce a difference (between which and witch for example) but I don't really perceive it (in my own or others' speech).
    That is, I perceive which and witch as homonyms but I definitely tend to produce them differently (I was pretty suprised to find this out). I won't say I produce the difference all the time but in citation forms the production difference is there.
    I assume this is a part of the process of loss of distinction between the two, I've known at least one person who did something similar with the cot/caught merger- she produced a difference at least in the citation forms but didn't really perceive it (as in she couldn't tell if I had the merger or not [I do]).

  28. Tim Martin said,

    April 10, 2011 @ 3:11 am

    Non-native, fluent Japanese speaker here.

    I feel like it's kind of superfluous to comment when Jim Breen has already done so – assuming this is the man who made the dictionary (thanks, btw!) – but for the record I agree with him. This is English forced into standard Japanese phonology.

    As for #3, I don't have experience with any other languages, but I don't find this phenomenon in Japanese all that extraordinary. Japanese people literally don't know how to, say, make a T sound without following it with some vowel, so Japanglish has to follow their phonological standards. "Deribarettori-" is just what you get when you run the word "deliberately" through a Japanese brain. No accomodations for English are made because none can be made.

  29. David said,

    April 10, 2011 @ 3:26 am

    In my ears, /hw/ is hwat most ever Scottish person says.

  30. Gunnar H said,

    April 10, 2011 @ 3:34 am

    Family Guy has a recurring joke about the hw pronunciation, the premise of which is that Stewie uses it and Brian finds it annoying.

    Personally I perceive it as slightly affected, probably because I was never exposed to it when I was first learning English.

  31. David said,

    April 10, 2011 @ 3:35 am

    Erm… I meant "every". Sunday mornings and typing don't go together for me.

  32. dw said,

    April 10, 2011 @ 5:08 am

    there is evidence that the presence or absence of /h-/ in words like whip and humor is often considered a test of social acceptability.

    Slightly off-topic, but in /hj/ onsets (e.g. "huge"), the sociological situation seems to be almost the reverse of /hw/. I never hear anyone drop the /h/ in real life, but pronouncing "huge" as /juːdʒ/ almost seems to be a job requirement to read the news on NPR. Perhaps it's an East Coast thing (I'm a Brit export to the US West Coast — dropping the /h/ in such words is very rare in Britain).

  33. Victor Mair said,

    April 10, 2011 @ 6:44 am

    I'm grateful for all the wonderful comments.

    As a frequent visitor to (and sometime resident of) Hong Kong, I was especially intrigued by Ron's comment about the English of Hong Kong doctors.

  34. Nathan Hopson said,

    April 10, 2011 @ 7:41 am

    FWIW, the English taught in Japanese schools was originally British English (the status English of the time). This has been almost, but not completely, replaced by American English (the status English of the present). The 'hwī pronunciation is a vestige from the era of British English.
    I haven't yet had enough coffee this morning to recall many others, but someone more alert will surely be able to come up with a few. The one that jumps to mind, though, is the first (and sometimes only) greeting series taught:
    A: "How do you do?"
    B: "Fine thank you. (And you?)"
    Presumably, this hangs around b/c of its perceived formality and syntactic match with:
    A: お元気ですか。
    B: はい、おかげさまで。(お元気ですか)

  35. Nightstallion said,

    April 10, 2011 @ 9:15 am

    In school (Austria), we were taught the h-less pronunciation of words starting with wh-, but I've since changed that part of my pronunciation just because I like the sound of it so much. ;)

  36. John Ward said,

    April 10, 2011 @ 9:33 am

    Hmm. Interesting stuff.

    It reminds me of the time I was trying to move a refrigerator in my apartment in Korea, and the Korean who was helping me move it said something about the "heels" being stuck. It took me a moment to figure out that she meant the "wheels." Apparently, initial hw is just as troublesome in Korean as in English, but instead of reducing the cluster to W, she systematically reduced it to just H, which was consistent with the local W-dropping Korean accent (this was in Chungcheong-Bukdo). She didn't realize that she dropped this sound, so her expression of the digraph was the exact opposite of most American's.

  37. Tenshi R. said,

    April 10, 2011 @ 10:03 am

    I feel like I heard a "th" sound accurately pronounced somewhere in the last third of the recording, when I would have expected an "sh" sound from a native Japanese speaker. That fricative is one of the most difficult for Japanese English students, so I'm confused about the speaker in this recording.

  38. Gregory Dyke said,

    April 10, 2011 @ 10:11 am

    I'm surprised noone else has commented on this, but I'm fairly certain that most foreign languages, spoken by people with little flexibility in their native phonemic maps come out just as standardised and regular. Possibly in a less obvious manner than japanglish, depending on the phonological constraints of the native language, the desire to pronounce correctly and the number of people with reasonable competence (in all but phonological aspects).

    Both English as spoken by a native French person and French spoken by a native English person are fairly regular and exhibit most of the phonological patterns of the native language.

    Typically in native french spoken english:
    – no word initial glottal stop
    – no dark l
    – /d/ and /z/ as possible allophones for /D/
    – /t/ and /s/ as possible allophones for /Z/
    etc. etc.

    Any counter examples?

  39. languageandhumor said,

    April 10, 2011 @ 10:12 am

    Further on the /hw/ front, the Japanese have borrowed the word "whistle" in the train sense. They write it ホイッスル (hoissuru) with HO + I) for an approximate /hwi/, not ウイッスル (uisseru) with U + I for an approximate /wi/ or the more recent ウィッスル (wisseru) with U + subscript I for /wi/. But they don't pronounce it ho-I-ssuru, which would imitate "whistle" with a /hw/. They pronounce it more like HOY-ssuru or HOYCE-uru, making it sound quite unlike like either pronunciation of English "whistle," which, of course, was never the target for a loanword.

    However, note that "whiskey" is ウイスキー (uisukii) not ホイスキー (hoisukii) even though it came into the language somewhat earlier (Meiji era, 1868–1912) than "whistle" (Showa era, 1926–1989), according to Kihon Gairaigo Jiten (Ishiwata, 1990). This could be from a difference in the English sources they heard or randomness.

  40. Andrew (not the same one) said,

    April 10, 2011 @ 1:29 pm

    dw: Might the pronunciation 'yuge' be a reaction against a possible 'hooge'?

  41. dw said,

    April 10, 2011 @ 2:47 pm

    dw: Might the pronunciation 'yuge' be a reaction against a possible 'hooge'?

    Perhaps — although I've never heard "hooge". The only accent I am aware of that might have "hooge" would be English East Anglian.

  42. Jongseong Park said,

    April 10, 2011 @ 3:24 pm

    John Ward, your story is somewhat surprising. Korean unquestionably uses a lot more initial /hw/ than English in its native and especially in its Sino-Korean vocabulary—Hwang is a common surname, for example. Initial /hw/ is retained in most traditional English loanwords, e.g. hwaiteu 화이트 (white) and the hwil 휠 (wheel) of your example, although 'why' would probably be transcribed wai 와이.

    In Korean, /w/ arises from what are phonologically rising diphthongs like |oa| or |ui|. There are dialect areas that lack such diphthongs, like the Southeast, so such speakers would have problems with /hw/. Former president Kim Young-sam who hails from this region was well known (and caricatured) for pronouncing hwaksil 확실 as if it were haksil 학실. But speakers from the Chungcheong region wouldn't have difficulties with such diphthongs.

    I think what may be causing the trouble is the realization of Korean /w/. This is a far weaker approximant than in English. The Korean /w/ is basically insertion of some lip-rounding at the beginning of the following vowel, and thus is subject to more assimilation to the following vowel. For instance, /wi/ is actually closer to [ɥi]. So when I pronounce the Korean hwil 휠 to myself and compare it with the English 'wheel', I could see how English speakers would not hear this as /w/. But Korean speakers, at least those not from the Southeast, definitely have /hw/ in these words.

  43. languageandhumor said,

    April 10, 2011 @ 5:18 pm

    @Andrew (not the same one) and dw:

    I think "hoo" is even further north than East Anglia. On Youtube, geriatric1927 (currently Derbyshire, but grew up nearby; Wikipedia says Leicester) says "human" as "hoo-man" (and "music" as "moo-sic").

    I think the [ç] voiceless palatal fricative allophone of /h/ in /hju/ (huge, human, humor) was selected out over time for a lot of dialects, usually just dropped to "yoo." This is unfortunate for language learners, as this is the same sound as in Japanese "hi" and "hyu" (and in German "ich"). Interestingly, the Japanese borrowed the word "humor" as R-less "you-more," ユーモア yuumoa when ヒューア hyuuma (British humou[r]) would be a nearly perfect equivalent. Yet when they use "human" in titles, it's not ユーマン yuuman but ヒューマン hyuuman.

  44. languageandhumor said,

    April 10, 2011 @ 5:21 pm

    Correction: "ヒューア" (*hyuua) above should have been "ヒューマ" hyuuma. I didn't hit the M hard enough in katakana mode.

  45. David said,

    April 10, 2011 @ 7:41 pm

    I grew up in Connecticut, and never pronounced or heard "wh" as anything different from "w". When ("wen") I first encountered people who pronounced the difference, I took it be an affectation. But I eventually came to realize that there are places (Alabama for instance, as shown in your map) where it really is different. My reaction was that it must be so much easier for children there to learn to spell! They don't have to memorize all those silent letters! However, I still have yet to meet anyone who says kuh-naif (knife). Anyway, whaling on the Pequod and wailing on the saxophone will always be the same to me.

  46. ug said,

    April 10, 2011 @ 9:13 pm

    I know a person who really emphasizes the "hw" sound. For some reason, it really annoys me to hear it. It's not at all like I think it's wrong, but I just don't like it. It sounds pretentious and usually the stuff that the person is actually saying doesn't match in register.

  47. LQ said,

    April 10, 2011 @ 9:47 pm

    "nowhere had she observed a person of true culture who did not possess that cluster"–more or less what English teachers in both North Carolina and Arkansas scolded me with, regarding "what" and "why." (I *am* a native speaker of English, but my father is Californian and my mother is Missourian/Californian.) I steadfastly refused to say it.

    Many Japanese guidebooks for the Anglosphere, as well as phrasebooks in general, use completely katakanized English to represent full English sentences (e.g. ハウ メイ アイ アシストュー? hau mei ai ashisutsyuu "How may I assist you?" or ウワッティズ ダ ゲイト ナンバァ uwattizu da geito nanbaa "What is the gate number?") I really think this is one reason why Japanese learners struggle so much with pronunciation. Since they've already learned 4 writing systems (katakana, hiragana, kanji, and romaji), one of the first things I usually do is teach them basic IPA.

  48. John Ward said,

    April 10, 2011 @ 11:06 pm

    Jongseong Park, I only said that this was in Chungcheong-Bukdo, not that it was typical of the Chungcheong accent. The Chungcheong accent in places like Daejeon was slightly different (clearer, more like Seoul) and I could clearly heard the labial approximate there. I was never able to find a good linguistic description of the accent in the hilly north-east where I was, but that's where the Chungcheong, Gangwon, and Gyeongsang accents mix, and my experience they do suppress the labial approximate. In fact, I remember being happy when my Korean teacher changed to one from Seoul because I could understand her better than the locals.

  49. WillSteed said,

    April 10, 2011 @ 11:43 pm

    Having learnt Japanese as a high school student and had intermittent connection with Japanese speakers since, and furthermore being quite familiar with the movie, I had little trouble understanding the passage.

    I heard "something" with dental fricatives a couple of times in there, and after the first few times, all of the 'why's were [waj], not [hwaj]. The transcriber (I'm assuming this was read off a katakana script), made good use of liaison between words to avoid word-final vowel epenthesis, e.g. for 'made it' meiditto instead of meido itto.

  50. Jongseong Park said,

    April 11, 2011 @ 3:39 am

    Thanks for your explanation, John Ward. Yes, I suppose it would be difficult to find descriptions of the accents in such boundary areas… I had hastily assumed what you heard would be something like the Daejeon accent, and missed the part where you said the local accent was W-dropping.

    The Chungcheong accent incidentally is supposed to be on the conservative side regarding vowels, so in the older speakers one may hear the monophthong /y/ (between [y] and [ʉ]) instead of /wi/ for 위. I don't know if this also applies to loanwords where /wi/ instead of /y/ is intended. Both /wi/ and /y/ in foreign words are adapted as 위 in Korean, but I don't know of anyone distinguishing the Korean pronunciation based on the source language; those who normally pronounce 위 as /wi/ use this pronunciation throughout even for cases where 위 represents /y/ in the source language.

  51. Robert said,

    April 11, 2011 @ 4:42 am

    The "safeword is whiskey" guy isn't correctly realizing the whine-wine distinction. He uses hw in weird and will, although he correctly uses ordinary w for way.

  52. chocolatepie said,

    April 11, 2011 @ 1:45 pm

    I'm going to propose a socio-historical account for the /hw/ in English loanwords. Loanwords really started to take off in Japan when English became a prestige language there, which was during the post-WWII occupation. This was the same period of time that /hw/ was fairly common among upper class white Americans. This socioeconomic group is where Japan might have been looking to model their pronunciation at the time, and once a loanword sticks, good luck changing the pronunciation.

  53. Victor Mair said,

    April 11, 2011 @ 8:21 pm

    From Bob Ramsey:

    Really an interesting discussion, Victor, and there are too many things that deserve comment even to begin to pick up the thread at this point.

    But one thing I might point out (and I may have missed someone's mention of this) is that the Japanese language in the 21st century is filled with what are sometimes called "innovative pronunciations", written with (fairly) newly devised katakana orthographic devices. These are mostly pronunciations modeled on English. One of those pronunciations that was thoroughly integrated into Japanese phonology a good while ago is テイー for 'tea'. Now, [ti] no longer exists in native vocabulary, as I'm sure you know (and because of that, 'team' became チーム). But these days almost no speakers under the age of 80 have any trouble talking about 'tea' quite naturally and easily. Similar katakana tactics take care of such things as the /hw/ you and the others have been talking about–and that the Japanese imagine is found in standard English, whether Brit or Yankee. My point is, the Japanese phonological system seems to be changing fairly rapidly, I think, under the influence of the sounds of English (or at least those represented overtly in writing). The Japanese famously do not distinguish [r] and [l], of course, but at least PHONETICALLY you hear [l] all the time now. And they're working on expanding the range of their bilabial fricative to accommodate that of English labiodental /f/. They haven't really got an equivalent of /v/ yet, but they've got a katakana spelling to help them try.

  54. Chris Davis said,

    April 11, 2011 @ 9:21 pm

    Some of the comments seem to suggest that the guy in this video is a passive victim of his native phonology, but I am almost certain that this whole thing is very meta.

    In my experience, bilingual Japanese speakers can very easily "switch on" this sort of pronunciation. And even when Japanese speakers with good English skills use an English word (for example, a place name) in a Japanese sentence, they almost always seem to fix the phonology, even if everyone in the conversation is bilingual, and even if the word in question does not exist as a loanword in Japanese.

    For example, I live in Amherst, and have many Japanese friends who speak English quite fluently, but everyone pronounces "Amherst Coffee" as アマーストコーヒー "amaasuto koohii", "Haigis Mall" as ヘーギスモール "heegisu mooru", etc., when speaking Japanese. Oh, and when speaking Japanese, you've gotta rephonologize your name — it's out of the question that people would call me anything other than "kurisu" when speaking Japanese.

  55. J. Goard said,

    April 11, 2011 @ 11:53 pm

    @John Ward, Jongseong Park:

    I don't know of anyone distinguishing the Korean pronunciation based on the source language; those who normally pronounce 위 as /wi/ use this pronunciation throughout even for cases where 위 represents /y/ in the source language.

    Indeed. I occasionally participated in a French class at my Korean university (this is Gwangju), and I'd say that a majority of the students were not only consistently pronoucing /wi/ for /y/, but unable to hear the difference reliably.

    My theoretical orientation surely might be biasing my observation here, but I can't help but suspect a frequency effect in the shortening of /w/ (or reduction to rounding on the consonant) that Jongseong describes. (See Joan Bybee (2003), Phonology and Language Use, particularly the discussion of pre-liquid schwa-dropping rate in words like every, memory, summary, mammary.) I'll bet that everyday words like 화장실 'bathroom' and 과자 'cookie, snack' tend to have a significantly weaker /w/ signal compared with phonologically similar but rarer words like 화제 'topic of conversation' or 과정 'course, curriculum'.

  56. Ethan said,

    April 11, 2011 @ 11:55 pm

    @Robert: That's intentional. It's a comedy.

  57. Nathan Myers said,

    April 12, 2011 @ 5:39 am

    What a fascinating post. I, too, wondered on first hearing whether the "hw" corresponded to any Japanese phoneme. Unlike most posting here, though, I am astonished to learn that aspirated "why" has fallen from fashion, and that anyone found the speaker's choice of aspiration odd. I'm certainly used to hearing unaspirated "wh" spoken by heathen savages, but had no inkling that it was considered standard anywhere.

    (For the record, I am from Hawaii, and have lived since mostly in Oregon and California.)

    In response to Prof. Mair's questions, I wonder whether there is a sharp divide between languages transcribed in syllabaries vs. alphabets. Surely syllabary languages have a smaller fund of familiar sounds to draw upon, resulting in more tortured adaptations?

  58. Jongseong Park said,

    April 12, 2011 @ 7:14 am

    J. Goard: I can't help but suspect a frequency effect in the shortening of /w/ (or reduction to rounding on the consonant) that Jongseong describes.

    You may be on to something. Even Koreans who might have trouble hearing the more subtle W-reduction (as opposed to complete W-dropping, which is dialectal) would recognize that a /wʌ/ following a consonant may reduce to [ʷo] in casual speech in cases like haejwo 해줘, which comes out a lot like haejo 해조. This might be because of the frequency of jwo 줘 (literally, the imperative 'give', which attaches to other verbs to form requests). I can't really think of parallel examples that resist reduction, though. In any case, it would be interesting to see an experiment or an analysis of sound archives to test this hypothesis.

    I just realized that the common erroneous substitution of yeokhwal 역활 instead of the correct yeokhal 역할 'role' could be a hypercorrection for W-reduction, although another probable factor is that the syllable hwal 활 is more common in familiar Sino-Korean vocabulary than hal 할.

  59. Ellen K. said,

    April 12, 2011 @ 8:49 am

    @Nathan Myers.

    Personally, I didn't find the speaker's use of hw odd. In fact, I didn't notice it at all. Now, if someone with my own accent did it, I'd probably notice it and find it odd. Not part of my idiolect, and I think not part of the local accent in the places I've lived. But listening to this Japanese person, I just heard the W phoneme, without noticing that it was the hw variation of it. I had to really listen closely, with the expectation of hearing it, to catch the h sound.

  60. Jeff said,

    April 12, 2011 @ 4:09 pm

    Languages that use syllabaries do so because they do not have a large number of sounds and they have a strict pattern of syllable construction. That would still be true even if they used an alphabet. Syllabaries are actually easier to learn to read than alphabets; since each letter is a complete syllable, the reader can simply read one after the other, instead of struggling to learn how to combine individual sounds into something that sounds right. So for a language that has a sound system well-suited to a syllabary, it makes sense to use one.

    At any rate, I don't think this example of Japanglish is any worse than what an English speaker would sound like reading romanized Japanese, so in that sense I don't think the alphabet really gives us much more of an advantage when it comes to approximating foreign words. French uses the same alphabet we do, but I guarantee you could find a French speaker who would produce an equally (if not more) tortured adaptation of this scene. Korean has an alphabet, but English loanwords in Korean are even more unrecognizable than in Japanese.

    I think it's hilarious how English learners of Japanese often complain about how loanwords are completely butchered in Japanese, as if English does a way better job. Even with the advantage of a more extensive sound system and many more possible syllables, not to mention a long history of absorbing words from many different languages, we are still able to mess up loanwords just as well as any other language. (Carry-okey anyone?). I don't see a problem with this, since it's not like the point of loanwords is to make it easier for speakers from that language to learn the language the words were borrowed into.

    As Bob Ramsey pointed out, katakana is adaptable and recently new combinations of letters that better represent foreign sounds have become common. Japanese had no 'fa' sound, but it can be written ファ by combining 'fu' with a small 'a.' All other non-native consonant+vowel combinations can be written the same way, since it's just an extension of the method already used for native combinations like 'shi' + 'yo' = 'sho' (しょ). The original loanword for "radio" is transcribed as "rajio" (ラジオ) but in some places I have seen it written "redio" (レディオ). 'Di' is not a native sound in Japanese, but it is becoming more common to hear it used. Consonant clusters present a bigger problem, but to some extent it's possible to approximate a cluster by using letters that end in 'u' or 'o'. So スbecomes 's' and ト becomes 't.'

  61. m.m. said,

    April 12, 2011 @ 7:09 pm

    dw said,
    April 10, 2011 @ 5:08 am
    Slightly off-topic, but in /hj/ onsets (e.g. "huge"), the sociological situation seems to be almost the reverse of /hw/. I never hear anyone drop the /h/ in real life, but pronouncing "huge" as /juːdʒ/ almost seems to be a job requirement to read the news on NPR. Perhaps it's an East Coast thing (I'm a Brit export to the US West Coast — dropping the /h/ in such words is very rare in Britain).

    Reading this, my mind goes directly to Robert Siegel, who happens to be from pre-[j] [h]-dropping new york city. Although, plenty of other personalities on NPR are older folk who were born closer to 50's, which would and does show in their english as more 'conservative' sounding. I'm curious how, if they exist, their post 80's whine-wine area born personalities align accent wise.

    chocolatepie said,
    April 11, 2011 @ 1:45 pm
    and once a loanword sticks, good luck changing the pronunciation.

    But they're saying it wrong and it sounds uneducated!


  62. Pflaumbaum said,

    April 13, 2011 @ 4:08 am

    @dw – /juːdʒ/ for huge is widespread in Yorkshire, especially the East Riding. Also parts of the Midlands, I think. In most other places the /hj/ it's pronounced [çj] or just [ç] – so perhaps it's on its way to becoming /j/?

  63. Jongseong Park said,

    April 13, 2011 @ 5:28 am

    @Jeff: English loanwords in Korean are even more unrecognizable than in Japanese.

    Curious which examples prompted you to say that. Most Koreans learning Japanese seem to think it's blatantly obvious that Japanese completely mangles English loanwords to a far worse extent than Korean, routinely bringing up examples like Ma-ku-do-na-ru-do for McDonald.

    Koreans are worse equipped in hearing the shortcomings of their own native phonology (the Korean Maek-do-nal-deu isn't miles better in my opinion), but Korean phonology certainly has a fair number of advantages over Japanese in mimicking the sounds of English, such as the existence of a lateral /l/ and the possibility to end syllables in stops /p, t, k/ and in distinct nasals /m, n, ŋ/. Korean also has a larger vowel inventory (having a separate vowel that can stand for /ə/ or /ʌ/ is particularly useful in transcribing English) which even allows a dedicated epenthetic vowel eu /ɯ/ so that one can tell immediately from its existence in a loanword that there is no vowel in the original.

    Japanese scores better in fricatives, as it has sounds that are better approximants for /f/ and /z/ (although /f/ is becoming common among Koreans as a marginal phoneme), and unlike Korean it can express vowel length distinctions in writing. Still, on balance, I say Korean

    All that said, English loanwords in Korean can suffer in terms of mimicking the sounds of the original language because of an insistence on one-to-one correspondance of phonemes over trying to mimic the different allophones ('star' is seuta /sɯtʰa/ instead of seutta /sɯt⁼a/ because the English 't' is always mapped as an aspirated consonant) and often because the original pronunciation was not known correctly. In fact, the most unrecognizable English loanwords in Korean tend to be those that came through Japanese.

  64. Sid Smith said,

    April 13, 2011 @ 9:07 am

    Brit here. Agreed that for us there's no h in what. I always associate it with the Katherine Hepburn era of American High Society films. Stewie in Family Guy seems to play to US stereotypes of Britishness – gay, affected, etc – and maybe the wHip pronunciation is part of that.

  65. Eric said,

    April 13, 2011 @ 11:36 am

    re: hwiskey, Love the Dudes, but I'm pretty sure the "Family Guy"/Stewie bit predates this movie, which makes it unfunnier.

  66. Dw said,

    April 13, 2011 @ 2:03 pm

    @Sid Smith:

    "Brits" do have the /h/ in "what", if they happen to be the Brits living in Scotland :)

  67. J. Goard said,

    April 14, 2011 @ 12:51 am

    @Jongseong Park:

    Most Koreans learning Japanese seem to think it's blatantly obvious that Japanese completely mangles English loanwords to a far worse extent than Korean

    With my native English, intermediate Korean, and beginning Japanese, it's blatantly obvious to me, too. Granted, I've lived with the Korean language, and Koreans' L2 English, for a while now, and only sporadically use my recently acquired Japanese, but I'm confident that the Korean adaptation of "McDonald('s)" would always have been more comprehensible to me and felt more accurate.

    It's interesting that you didn't mention what for me is probably the single biggest factor affecting comprehensibility, namely the timing system of Korean (somewhere between syllable- and stress-timed) versus the moraic timing of Japanese. Properly contracting sequences of sounds, whether it's word-internally, within highly entrenched expressions, or merely in backgrounded information, is a huge part of sounding natural to native English speakers, and L1 Korean speakers sure seem to have a native advantage over Japanese in that regard.

    That said, I think the key here is that Jeff referred not to direct adaptation of English to Korean, but to "English loanwords" in general, many of which have been filtered through Japanese phonology on their way to Korean. If this is indeed the source of his comment, then it should come as no surprise! Like the "telephone game", it doesn't take many steps before a signal is completely unrecognizable…

  68. Ethan said,

    April 14, 2011 @ 1:23 am

    And like the telephone game, the circle can be completed when the loan word is borrowed back unrecognized into its originating language. Jeff points to "karaoke" as being a Japanese word badly mangled upon import into English. But "karaoke" itself was itself half-borrowed from English, as the "oke" part is a typical Japanese borrow-then-derhotacize-then-shorten version of "orchestra".

  69. Denis said,

    April 14, 2011 @ 8:09 am

    For not-very-fluent-English-speaking Russians, it is pretty common to try pronounce all letters in English words if there is a need to spell an English word (especially in technical areas heavily dependent on using English terminology, e.g. software development). This, of course, produces an extremely heavy Russian accent but solves the problem of making the spelling across the conversation. Even fluent English speakers might do it every now and then, especially if they are not sure on the other sides' knowledge of the English word's spelling.

  70. John Ward said,

    April 14, 2011 @ 8:58 am

    @ J Goard, Jongseong Park

    If I had to hazard a guess at what was happening, I'd say that the affected vowels are underlying rounded-front vowels that manifest differently in different regions. In fact, in Jecheon, what I think was happening when I said that the labial approximant was dropped was really just that the rounded-front vowels were unrounded after a consonant. That is, lip rounding become a feature of the initial consonant rather than the nuclear vowel, and most of those rounded consonants later merged with their unrounded counterparts. Incidentally, the two types of vowel were probably still distinct in at least some cases as far as native speakers are concerned. I'm not sure about all the rounded-front vowels, but I do clearly remember from one lesson that the difference between 채 and 최 was just that the vowel was more central in the latter case. So something similar could have applied to the other vowels, even though I had trouble hearing it. That is, backness could have replaces roundness, which has a similar effect on the formant frequencies.

    So it looks like there are four possible ways for underlying rounded-front vowels to manifest in Korean: rounded-fronts, rising diphthongs, centralized vowels, or simply merged with the unrounded-fronts.

  71. Jongseong Park said,

    April 14, 2011 @ 11:34 am

    @ John Ward, my own usual pronunciation of choe 최 is something like [tʃʰʷɵ] or [tʃʰʷɘ] (the rounding is there, but it's not strong). Like most young Koreans, I usually use rising diphthongs for the rounded-front vowels, but after affricates these frequently become partially rounded central (or mid-front at most) vowels, with the preceding consonant acquiring the lip rounding. By contrast, chae 채 is between [tɕʰɛ] and [tɕʰe] for me, depending on how lazy I am about maintaining the height difference between ae and e. The rounded-front vowels, at least in the limited distribution they have in my idiolect, are definitely more central than the usual transcriptions as /y/ and /ø/ would suggest; I would have chosen /ʉ/ and /ɵ/. My rising diphthongs /wi/ and /we/ are not noticeably centralized.

    In terms of perception, I would guess the most salient difference between choe and chae as produced by speakers like me (I'm from Seoul) is the consonant quality. In citation form, I can produce a five-way distinction between choe 최, chwe 췌, chwae 쵀, che 체, and chae 채. These distinctions would surely be muddled in casual speech, but I would much more easily confuse choe, chwe, and chwae with each other or che and chae with each other due to vowel-merging tendencies than confuse a syllable that starts with a phonetically rounded consonant with one that doesn't.

    In my experience, though, W-dropping Gyeongsang-do accent speakers use the same consonant quality in all of the above examples. So between the two extremes, it wouldn't be surprising if there were some speakers who distinguished choe and chae solely through vowel quality. It may depend on the consonant, too; for me, the biggest difference between goe 괴 and ge 게 seems indeed to be that the former vowel is more centralized. I'm sure there is huge variation among individual speakers.

  72. Victor Mair said,

    April 14, 2011 @ 9:11 pm

    From a correspondent: Your post says, "The katakana transcription of 'why' is ホワイ(ho wa i); 'where', ホエア (ho e a) or ホエール (ho ē ru), and 'what' is ホワット (ho wa t to)."

    I think that ホエール (ho ē ru) is usually a transcription of 'whale', not 'where'.

  73. Jeff said,

    April 15, 2011 @ 2:12 pm

    Ok, I think I must have phrased something wrong, because everyone misunderstood what I was saying about loanwords. What I meant to say is that no language is "better" than any other language at adapting foreign words into their own vocabulary. Loanwords are not supposed to be phonetically accurate representations of the original word; they are supposed to blend in naturally with the native sound system.

    English has twice as many vowels as Spanish, but they don't match up, so it doesn't give us an advantage in pronouncing Spanish loanwords the "right" way. I'm a native Spanish speaker, but if I have to say a Spanish word or a Spanish loanword in an English sentence, I pronounce it with an English accent. When I speak Spanish and I say a word that was originally Arabic, I say it with a Spanish accent.

    Do you guys really think it would be more correct for everyone to pronounce loanwords exactly as they were pronounced in the original language? Personally I think it would sound far more unpleasant if Japanese was peppered with properly pronounced English/Portuguese/German words. English would also be a mess if we did that, considering all the different languages we've borrowed from.

  74. Jongseong Park said,

    April 15, 2011 @ 6:26 pm

    Jeff, the only thing I was responding to in your comment was your curious assertion that "English loanwords in Korean are even more unrecognizable than in Japanese." (I'm still wondering what examples prompted that.) I've skimmed through the comments again and I don't think anyone here is claiming that one should pronounce loanwords exactly as in the original language. I don't think anyone with enough linguistic knowledge would make any such claim.

    What I meant to say is that no language is "better" than any other language at adapting foreign words into their own vocabulary.

    In general this is true, but if we are talking about a specific source language, then some languages definitely can be better than others at approximating the original pronunciations due to phonologies that happen to correspond better. This is what some of us were discussing regarding the merits of Korean vs Japanese in approximating English. I would also think, for example, that Greek would be better than English at approximating Spanish loanwords. (There are many other aspects of word borrowing that we could deal with, of course, but in this discussion we are mostly talking about adapting foreign sounds to native phonological constraints.)

    I agree with Chris Davis above that the clip which is the subject of this post seems to be very meta. Of course we adapt foreign sounds in loanwords to our native sound systems all the time, and those of us who know the original pronunciations are well aware of the distortions involved. This clip I think is an example of someone having fun with that, applying these distortions to an entire soliloquy rather than to isolated loanwords to heighten the effect. One could just as well imagine someone repeating the dialogue from a Japanese film but using in a totally unadjusted accent of an English speaker, and results would be similar. I don't think anyone here finds anything wrong with the fact that there are such distortions in pronunciation.

  75. Richard said,

    April 18, 2011 @ 6:58 am

    It always feels awkward for me to say 'karaoke' to non-Japanese people. I know if I pronounce it the correct way, they won't know what I'm on about. I usually end up with something intermediate between the Japanese pronunciation and typical English one. I reckon the word 'sake' would be better pronounced by English speakers (at least, Aussie English speakers) if it were romanised as 'sukkeh'.

  76. Taylor B said,

    September 2, 2012 @ 5:43 pm

    Interesting questions and comments here but I am little unclear about a few things. I'm assuming that what we are doing is a phonological assessment of this speaker relative to two phonological systems (English and Japanese), the question being which phonological features of which language is this speaker conforming to. The most relevant feature being phonotactics, and then prosody (a subject I don't know enough about to do a job proper here I apologize).

    The hypothesis is that this speaker is reading a kana transcript as opposed to be doing real time conversion (Eng to EngJap). To be sure, this speaker could have memorized and rehearsed a transcript, with no actual reading involved in this video. Since Kana is almost in perfect correspondence with segmental features and encodes only mora (the mora is the smallest prosodic domain in Japanese, the foot being the second smallest), if there is a transcript involved then all segmental coding has been done by the one who wrote the transcript. This speaker, the reader then, is charged with all prosodic decisions above the mora, and has not had to make any decisions at the segment level.

    In either case, the surface segments seem to be in conformance with Japanese phonotactics. Specifically and expectedly, phonotactic conformance seems to be with the loanword strata (外来語) of the Japanese phonological lexicon, and not the more stringent miniphonology of the native stratum (和語). A simple set of rules (google Japanese Loanword Phonology) can account for the segmental changes of loanword importation, and a segmental description of this particular speech can be treated as loanword phonology.

    But, from what I can gather with only my ears, it really does seem that the speaker's intonation is that of English and not Japanese. I would say that this speech act is segmental English altered in a way that minimizes Japanese phonotactic illicitness, but remains fully English in prosody, but this deserves greater elaboration than I can give!

RSS feed for comments on this post