Mangling Foreign Dialects

I've found the most amazing book. “Amazing” is a word that here means “fascinating and appalling.” It's called Foreign Dialects: A Manual for Actors, Directors, and Writers, by Lewis and Marguerite Herman, published in 1943. Remarkably, it seems to be still in print.

Don't try this if you're not him


Let's just dive in. Here is what is presented as a typical Russian speech. The authors proudly note that it was presented on television.

Oh! I very good fellow! Why? Because I Cossack. I very big Cossack. Yah! I captain of Royal Cossack Guard in Moscow - in old country. Oh! I got fifty - hundred - five hundred Cossack they was under me. I be big mans. And womens, they love me lots. Nastia Alexanderovna— she big ballet dancer in Czar ballet— Countess Irina Balushkovna, she love me. And men? Ach! They be 'fraid from me. They hating me. Why? Because I big Cossack.
And that's just the written text. They helpfully provide the pronunciation:
AW! I vAri gOOtTH falAW! vwI? bikAWs I KAWsAHk. AW! I vAri bEEk kAWsAHk. yAH! I kAHptTHAHn AWf rAWiyAHl kAWsAHk gOrTHs yEEn mAWskvAH— yEEn AWl' kAWntTHri. AW! I gAWtTh fEEftTHyi - khAWndTHAWrtTH — fIf khAWndTHAWrtTH kAWsAHk dTHEH AWndTHAWr mi. I bi bEEg mAHns. AHn' vwEEmAHns, dTHEH lAWf mi lAWtTHs. nAHstTHyAH AHlyEhksAHndTHAWrAWvnyAH - shi bEEg bAHlyEH dTHAHnsAWr yEEn tsAHr bAHlyEH - kAH:ntTHas yEErEEnyAH bAHlyooshkAWvnyAH, shi lAWf mi. AHn' mAn? AHkh! dTHEH bi 'frEHtTH frAWm mi. DTHEH khEHtTHyEEng mi. vwI? bikAWs I bEEk kAWsAHk.
They do provide an IPA key, so if you want to reproduce this speech, try this:
ɔ / aj værɪ gotð fælɔ! / vwaj? / bɪkɔs aj kɔsak / ɔ / aj værɪ bik kɔsak / ja! / aj kaptðan ɔf rɔijal kɔsak garðs jin mɔskva / jin ɔl kɔntðri / ɔ! / ai gɔtð fiftðyɪ - xɔndðɔrtð - fajf xɔndðɔrtð kɔsak dðɛ ɔndðɔr mɪ / aj bɪ big mans / an vwimans dðɛ lɔf mɪ lɔtðs / nastðja aljɛksandðɔrɔvnja ʃɪ big balyɛ dðɔnsɔr jin tsar balyɛ / ka:ntðæs jirinya balyuʃkɔvnja ʃɪ lɔf mɪ / an mæn? ɔx! / dðɛ bɪ frɛtð frɔm mɪ! / dðɛ xɛtðiŋ mɪ / vwaj? / bɪkɔs aj bik kɔsak
Wait, [tð]? Not a typo; they maintain that the Russian dental t “thickens the t sound so that it is pronounced with the voiced TH, as in tTHyi (tea)”. Now, you are actually well on the way to a good Russian accent if you pronounce [t d n s] as dental, but I assure you that dental [t̪] is not [tð] and doesn't sound like it.

Another WTF: a Russian might sprinkle palatalized consonants into his English, but even a beeg Cossack won't put it on Russian words that aren't palatalized: it's Irina, not Irinya. (Also, it looks like they were aiming at a surname, so that should just be Balushkova.)

Make the bambino

Here is an Italian… not a full Cossack Italian, just an immigrant:

My wife, she make the bambino and the doc he tell me to wait right here. Is all right with you? I no can sit still. I no can stand still. Oh! I no afraid, oh no! Is just I kind of nervous, that's all. Is not I think anything she happen to my Rosa. She so strong like the horse. She no got no operation— she no sick all her life— you bet my life, no! Say! What's the most kids born? the boy or the girl? I want the boy. You bet my life! Rosa, she want the girl. Son of a gun! What if she be twin?
At least they don't add -a to all the words, right? Oh dear, wait, look at the phonetic representation:
mI:uh wI:fuh shEE mEH:kuh duh bAHmbEEnAW AHnuh dUH DAW:kuh, hEE tAluh mEE tOO wEH:duh rI:d 'EE:ruh. Eez OH:luh RI:duh wEE:tuh yOO:?
They actually advise, “Many Italians prefer their own grAHtsEE-EH to the American 'thank you'.” There it is in black and white, the notion that foreigners can learn any word of English except for the handful of words Americans happen to know.

Oh, and here is how they think spaghetti is pronounced: [spagæ:də].

ROOlz fAWr fuhREHnsh

There's a whole chapter for each “dialect”, which begins with a catalogue of stereotypes. E.g. for France:
The average Frenchman is, first of all, a completely rational personality. He thinks first and then acts. But he does not think too much. He has too much love of life for sedentary thinking. That is why his literature, his plays and his philosophical works are witty and even frivolous.
One wonders if the authors were admiring or reproving here:
His system of morals and ethics does not jibe with the world at large. He accepts the Negro, the Oriental and all races, creeds, and colors as unreservedly as he accepts his fellow Caucasians. …A man's place in France is determined by his intelligence, his capabilities and his civilized attitude to life rather than by his behavior or the color of his skin.
Anyway, on to the dialect! First, instruction is needed for the troublesome French nasal:
To learn how to reproduce this quality, raise the back of the tongue until the throat passage is closed and the sound will be forced up through the nose. Do not hold the nose because the resulting sound will in no way be similar to a French nasal.
Can't fault that last sentence— you definitely don't make a sound characterized by airflow through the nose by closing the nasal passageway— but the first sentence is basically instructions to choke yourself. Although it's true that the tongue rises for a nasal, there is a far easier path. We have nasals in English: n, m, ng. Pronounce an extended n, then try to make that sound without touching the tongue to the palate. Voilà, a nasal airflow.

On [ə]:

This mute sound of “a” is changed by many Frenchman into AH, as in SAWfAH (sofa). …A variant for this “a” is the “oo” of “good”, as in sAWfoo (sofa).
Now the first part I can sort of see: if a Frenchman pronounces sofa in French, he'd say [sɔfa] (which actually exists as a borrowing). But the English shwa has a close analog in French— the vowel of le and emphasized final -e. Hearing English sofa, I would think the Frenchman would say [sofœ]. Maybe this is what the otherwise bizarre second part (sAWfoo = [sɔfʊ]) is trying to say.

Re the actual [ʊ], as in put:

The French usually give this short "oo" the "ER" treatment. The "ER" sound is present in the American word "curb" when it is pronounced without the "r".
They go on to suggest "pERt for "put".

The problem here is that there is no separate vowel in "curb". It's just a syllabic r: [kr̩b]. Some dictionaries write things like kûrb, but that seems to derive from a feeling that [r] can't really be the vowel. But it is, so the advice to leave out the "r" is impossible.

Fortunately the authors provide an IPA key, from which we can see that they think "curb" is [kə:rb]— not just with a vowel, but a lengthened one. Well then, if they mean [ə] they should say so. But I doubt a Frenchman would say [pət] for "put"; more likely [pœt].

Although a great many French people sound “g” so far forward in the mouth that it often resembles weak “k”…
Wut? G changes to k when you stop voicing it, not when you move the point of articulation forward.
Medially, “s” is pronounced as “z” when it comes between two vowels, as in pAWzEEbluh (possible)… Finally, “s” is always pronounced as “s”, even when it is sounded as “z” in American, as in rAWs (rose) [and] Amyüs (amuse).
Not buying this. At a first approximation, French has exactly the same rule for intervocalic s as English: <ss> = [s], <s> = [z]. There are exceptions, but French possible has [s], while rose and amuser have [z].
Ordinarily, this “zh” sound, found in such words as “azure”, …is treated as “zy” by the French[:] AHzyür
If they're reading the English words as if they were French, yes. But it's not like English [ʒ] is any mystery; it's just French j.

They helpfully provide a list of “typical expressions and interjections”: eh!, eh bien!, zut!, oui, non, n'est-ce pas, chéri(e), chic, c'est la vie, bien!, madame, mademoiselle, monsieur. Ze readers, it is très chic to sound très français, c'est la vie zut alors!


The first chapter I turned to was the one on Chinese. As with the French, they start with a portrait:
The Chinese are reserved, calm and even-tempered among foreigners. But in the sanctity of their homes… they laugh and titter and scold and are more spirited than is usually believed….

Whatever Occidental customs are filtering into Chinese life, there is one Oriental custom which can never be dimmed. That is the ceremonial— any ceremonial, all ceremonials. There is a joy beyond all joys during a festival. The colorful costumes, the frightening masks, the unrestrained leapings…

And, on the whole, they are a rather wonderful people. They seem to survive although their food is poor and scarce. Their spirit will not be broken although it be subjected to the cruelest of tortures and an interminable series of disheartening conditions. There is a proverb for every occasion and a wide grin to accompany it.

A little condescending, but generally positive… unlike the portrait of the Japanese:
He bows obsequiously, hisses politely and does everything he can to humble himself in the eyes of the person he is trying to impress. This almost insidious habit is, no doubt, a form of compensation with which he tries to overcome his strong inferiority complex. He is sensitive about his short stature, his color, his buck-teeth, his weak eyes and his country's comparatively recent emergence as a civilized nation.
Well, we were at war with Japan at the time. But back to Chinese! The Hermans know that there are several dialects, though listing both “Cantonese” and “Kwangtung” is a bad start. They then forget this for the rest of the chapter, never seeming to realize that most Chinese-Americans they would have met were of Cantonese origin.
The Chinese sing-song can be attributed to the fact that, in his own language, the Chinese has only four tones. …A Chinese would be a poor subject for the dialect coach who required that a student give the word "yes" some twenty different interpretations. He, being Chinese, would stop at four, achieving only a question, disbelief, a curt response and a whine.
Wow. First, and again, the Chinese-Americans that the Hermans knew would have mostly spoken Cantonese, which has nine tones. (Or six if you think the tones in checked syllables are allotones.) Second, even for a Mandarin speaker, it seems bizarre that they'd speak English with only the four Mandarin tones. Third, this is a weird confusion of lexical tone and intonation. Of course Chinese people can act, and pronounce the same word differently based on their attitude, purpose, and emotion.
The Chinese dialect changes “EE” to “i”, although for a lighter dialect the EE may be retained.
Wut? By “i” they mean [ɪ]; they suggest [mɪ] for me. In fact Chinese mi is precisely [mi].
When broken down, this American “I” becomes “AH-EE”. The “AH” is retained by the Chinese, but the “EE” is changed to short “i”. …The two sounds should be pronounced distinctly and separately.
So they give ice as AHiss, that is, [aɪs]. For other dialects the Hermans say all sorts of strange things about /aj/— e.g. for French they write AH-EE, with the last part emphasized. Jerry Norman reports that Mandarin /ai/ is [ae], which I guess is within spitting distance of [aɪ], but this is a strange case of over-articulation.
Initial “g” should remain the same as in American, although some Chinese pronounce it as “w”.
In words which end with ng, as “young” and “thing”, the “g”is dropped completely and the “n” is sounded as the final consonant, as in yAHn (young) and ssin (thing).
Wut? I can't imagine where this comes from; final ŋ is easy for Chinese.
The “j” also is more forced than it is in American. There should be a slight ”d” sound before the ”j” making it actually “dj”. If a thicker dialect is needed, the “ss” may be used.
And this when their own IPA chart tells them that English j is [dʒ]. Then comes the bonkers advice about a “thicker” dialect. Hey, kids, if you want to sound really racist, here's what you do!
A number of Chinese have difficulty with [k] and substitute a strong “h” for it. This “h” substitution may be used for a thicker dialect.
Oy. Cantonese and Mandarin both have [k]. They also want you to pronounce [kw] as “a grating sound”, apparently [x] or [χ]. Again, both Chinese dialects have their own [kw].
Although the Chinese could feel superior to the Japanese in that the Chinese can pronounce “l”, they don't seem to appreciate their speech advantage for they are very negligent with this consonant. So unless the “l” is absolutely essential to the word, it should be dropped….
From their examples, what they mean is that syllable-final l is dropped. They make everything much harder by refusing to teach some simple phonetic terms (like initial, final), to say nothing of using their half-dictionary half-phrasebook, Klingonish transliteration.
V - This letter often is sounded like “bv”. The lips are slightly parted and an attempt is made to sound both consonants at the same time. The lips do not close but remain half open until the sound is completed.
It can be only an attempt as you can't both voice and unvoice a consonant simultaneously. I don't know what they're getting at— perhaps [β]? If I were Chinese (and not from Shanghai) I'd probably go with [f].
SH - This sound also receives the hissed “ss” treatment.
Didn't they ever wonder about the name Shanghai? This is a Cantonese thing.

Also, “hissed”? They explain that s is pronounced with the lips as for English sh, which would mean slightly rounded. I don't find any indication that Mandarin or Cantonese [s] is rounded, but Cantonese [s]/[ʃ] is allophonic, perhaps that's where they got this idea…

Another typical word habit derived from their original language is the coupling together of the same verbs, as in “mEHkuh ssituh ssit” (make sit-sit) which actually means “sit down”.
Probably actually a representation of Chinese Pidgin English. I don't know how reduplication is used in Cantonese; in Mandarin, zuò-zuò would mean “sit for a while”.

Oh, want a monolog for your audition? Of course you do. The Hermans specify that their character should be “grinning widely”.

Ticket, please? Thank you. Me got wash finish. Tickey say, “Mr. Harris.” That you, no? Fifty-five cent please. Thank you. Maybe you put change in China Relief Box, no? That for China people Japan make hurt.

tikEH puhliss. ssEHnuhkuh yoo. mi gO wOssuh finiss. tikEH sEH, missuh hEHliss. dEHt yoo, nAW? fiti fAHi ssEHnuh puhliss. ssEHnuhkyoo. mEHbi yoo poo tssEHnuhss inuh tsAHinuh lilifuh bAWkuhs, nAW? dEHt fAW tssAHinuh pipoo tssAHpEHnuh mEHkuh hERt.

[tɪkɛ pəlɪs. sɛnəkə yʊ. mɪ ga wassə fɪnɪs. tɪkɛ sɛ, mɪsə hɛlɪs. dɛt yʊ, nɔ? fɪtɪ faɪ sɛnə pəlɪs. sɛnəkyʊ. mɛbɪ yʊ pʊ tsɛnəs ɪnə tsaɪnə lɪlɪfə bɔkəs, nɔ? dɛt fɔ tsaɪnə pɪpʊ tsapɛnə mɛkə hət]

If this is a Cantonese person, they can absolutely say [tɪkɛt], [tsapɛn], [mɛk]. (And if they speak Mandarin, they can say [tʃajnə].)

Who was this for?

Reading this book is like looking into an alternate timeline. I think if an writer wrote these speeches today, or an actor performed them as described, they'd be run out of town. Or they should be. Amusing as it might be to hear Vladimir Putin saying “I be big mans”, we prefer a good deal more subtlety today. Plus writing like this makes the character look like an idiot. (To be fair to the Hermans, Americans probably ran into strong European accents far more in their day.)

The authors claim to have interrogated actual specimen speakers of each dialect; still, their dialogs and presentations sound like exaggerated stage versions. At the same time, the book is a reminder that “diversity” isn't a new thing in America— it's just that the particular diversities are different. Today people can honestly feel that “white American” is a single class (very possibly divided into northern and southern); but to the Hermans' public, “white” was divided into English, Scottish, Irish, German, Russian, Italian, French, Polish, Yiddish… there are even sections on how to properly capture Lithuanians, Yugoslavs, Czechs, Finns, Hungarians, and Greeks. (Greek for “He made her mad”: [xɪ mɛkəm xʊɔr madð].)

The Hermans never explain their expertise, but since their dialogs are attributed to themselves, I assume they're screenwriters. They know a little linguistics (enough to make their IPA chart), but not enough.

I have to emphasize that these are pretty extensive chapters, with exercises for the reader. Actors had to study hard to make these horrible renditions. Before Youtube, at a time when buttonholing a random Lithuanian was tedious work, this book must have been a godsend to the vaudevillian and the character actor.

What do I do instead?

If you're a director, for God's sake don't have a white actor learn a Chinese accent, hire an Asian actor. If you need an English character, it's probably not that hard to find an English actor. I've seen reviews which complain about an actor using a wrong accent, but never one that complained that their accent wasn't thick enough.

If you're an actor, the general principle is to use the person's native phonology. English th is usually a problem; our wide array of lax vowels and consonant clusters may be, too. Substitute the closest sound; see my LCK to know what sounds actually are close. Don't overdo it; a little variation goes a long way. Err on the side of assuming a character can pronounce something rather than that they can't.

You can model your dialect on a native speaker, but try to understand the phonology behind what you're hearing— lest you hear things that really aren't there, like the Hermans' [vw] and [tð] in Russian.

Decide if your character learned his English (mostly) from reading or from listening. E.g. a French person who learned from books might pronounce pleasure as [plɛˈzyr], but one who learned from the spoken language could easily say [ˈplɛʒœr].

If you're a writer, don't include token foreign words (grazie!) and don't try to indicate the phonetics at all. Avoid anything that sounds like pidgin. Use syntax to indicate how the character speaks. Art Spiegelman's Maus is a great example: he shows off his father's Yiddish dialect entirely with word order.