Non-native pronunciations of English

Non-native pronunciations of English result from the common linguistic phenomenon in which non-native users of any language tend to carry the intonation, phonological processes, and pronunciation rules from their mother tongue into their English speech. They may also create innovative pronunciations for English sounds not found in the speaker's first language.

Grammar differences (for example the lack or surplus of tense, number, gender etc.) in different languages often lead to grammatical mistakes that are telltale signs of their native language. Sometimes non-verbal body language, facial expressions, or other cues also give away the origin of the speaker.

Another factor is how the English language is taught as a second language. In such cases, the accent of the teachers will affect the accent of the students. In some places that were formerly under British rule, such as India, Hong Kong and Malaysia, the English language remains a mandatory subject in most schools. Students study the English language at schools while using their native language in daily life. Thus, there may be distinctive features of pronunciation in those speakers.

A reverse version of this article is to be found at: Anglophone pronunciation of foreign languages

Non-native accents by region in alphabetical order:

Afrikaans

Afrikaans has a voiced rather than English’s voiceless , this may carry over when speaking English.
Non-rhotic accent. Typically, r may be otherwise trilled.
are unaspirated, which may sound like
Final obstruent devoicing.
Interdental fricatives and become labiodental and respectively.

Arabic

Trouble with many English vowels, since Classical Arabic has only long and short forms of three vowels: . Depending on the speaker's original dialect, the ease of learning English vowels may be facilitated by allophones of these three vowels.
is often pronounced like b - Arabic doesn't have .
is sometimes pronounced like f, since Arabic doesn't have .
The sounds , and may be confused, as they are dialectual variations of the same allophone:

Levant Arabic pronounces it as
Egyptian Arabic retains the archaic
Almost all other dialects pronounce the sound as
- Final obstruent devoicing.
is trilled, as it is in Arabic.

Bosnian, Croatian, and Serbian

Note: Classification of these dialects remains controversial. See Serbo-Croatian language and Štokavian for more information.

Often a palatalized dental is used instead of the typical English 'r', especially before vowels.
is pronounced as in Spanish José.
Difficulty distinguishing between and because the former does not exist in these languages.
Dental fricatives are typically replaced by apical stops.
Difficulty with English diphthongs. May pronounce each vowel as distinct.
Difficulty with gender pronouns, because personal pronouns may be used for normally gender-neutral inanimate objects as well (like in English) but with a different gender than in English. For example, English ship is usually neuter but can sometimes be feminine while it is masculine in Štokavian).
Difficulty with articles such as "a" and "the" as they do not exist in Štokavian languages; the closest approximations are demonstrative adjectives like "some" or "one" and "that".

Bulgarian

as in "the" or "then" is often pronounced as .
as in "thong" or "thorough" may be pronounced as .
Sometimes "he" or "she" is used where "it" should; on the other hand, ships could be referred to as "it."
Overuse of definite articles, especially in phrases whose Bulgarian equivalents require articles, like "the Bulgarians usually..." instead of "Bulgarians usually..."
Pronunciation of English as the velar fricative (as in Spanish Jose).
Voiced consonants at the end of the words may be pronounced as voiceless, like "bat" instead of "bad."
Tendency to pronounce as in words like "sun", "up", "under."
may be pronounced as an alveolar trill, absent in English, or with a hard rhotic accent at the end of words (like "car", "fire" etc.)

Cantonese (Chinese)

see also Hong Kong English

is often pronounced as or sometimes . Speech is usually non-rhotic (more common if schooled in Hong Kong because of British influence). and are also often confused since these two sounds are becoming allophones in Cantonese.
, often pronounced as either or .
'wh', pronounced as (this is the case for many native speakers of English as well).
Dental fricatives usually become labiodental.
Devoicing of final , and , making 'nip'/'nib', 'rot'/'rod' and 'flock'/'flog' indistinguishable.
Tendency to epenthesize after word-final postalveolars.
Final may become so that win/wing and ran/rang become indistinguishable.
Often dropping or not distinguishing articles like 'the' and 'a'.
Difficulty with verb tenses and plurals in general, as they have no direct equivalence in Chinese grammar.
Speakers may have difficulty with pronoun gender: (he, she, and it) since spoken Chinese does not make gender distinction in pronouns (although these words have different hànzì).
Marked conversion of stress into tones since Cantonese is a tonal language.
Introducing end-of-sentence particles that are non-existent in English (this occurs primarily in informal writing rather than in speech).

Czech

Devoicing of word-final voiced consonants.
Trouble distinguishing between and .
- With the two above combined, "bet," "bad," "bed," and "bat" will all sound like .
Frequent voicing of between vowels, and especially after , , and , so that "insert" and "increasing" sound more like "inzert" and "increazing" respectively.
The vowel in words like "bird" is frequently pronounced as or .
Prevalent mispronouncing of the 'u' in words such as "purple" or "Murphy" as a : the 'u' in these words is mistakenly believed to represent about the same sound as in "but", and are often lengthened.
(as in "heart") is murmured, as it is in Czech.
Word-final (as in "bill") is not velarized, as it is for a number of native English speakers, making it of the same quality as the initial 'l' in "lip."
Frequent lack of aspiration on unvoiced consonants, making "park" and "bark" harder to distinguish.
Difficulty in distinguishing between and since the latter does not appear in Czech phoneme inventory.
Prevalent misconception of 'oo' as representing a long vowel where it should be short (thus rhyming words as "book", "hook" and "foot" with "loot" instead of "put"). This is an example of spelling pronunciation.
'ng' is always treated as two separate sounds, as the sound does not exist in Czech; this, coupled with word-final devoicing, makes -ing frequently pronounced as "ink", making "thing" and "think" homophones.
Some trouble with the dental fricatives and , which may be rendered either as and or and respectively.

Dutch

Pronouncing voiced consonants as voiceless (/d/-> /v/ -> ^* etc), especially at the end of the words; for example, "pod" would sound like "pot."
Dutch does not have dental fricatives. Most commonly, becomes ("though" sounds like "dough") and becomes or ("think" sounds like "sink" or "tink").
Stereotypically, a Dutch speaker from around Amsterdam may make /s/ sound more like .
Speakers may pronounce either slightly rolling, as in Dutch, or hypercorrecting it which may sound more like 'w'.
The pronunciation of vowels differs widely across Dutch dialects. All dialects lack the in "putt", the in "pot" (although many American English speakers lack the second sound as well) and the in "pat." The latter of the three is usually realized as .
Speakers from Limburg can often be identified by their intonation, even when speaking English or most any other foreign language. Their native dialect has various tones, and characteristically sentences end in the up-and-down intonation.

Finnish

An absence of the English tonal variation. In Finnish the intonation contour is typically falling, and there's no meaning-distinguishing variation at the sentence level. Thus, speakers often have difficulties producing the rising tone of a question. The tempo of general speech is usually slower, which is how Finns typically speak.
There may be an absence of the word "please". In Finnish, politeness generally comes from tone of speech and the use of formal pronouns. The expression "ole/olkaa hyvä" (for "please") is generally used only in formal settings or when stressing a specific issue politely.
In Finland, the traditional teaching policy focuses on writing with correct spelling and grammar, but pronunciation or usage gets less attention. This may lead to pronunciations imitating the spellings, e.g. "font" pronounced as , unlike the native English ( in American English), or "count" becomes instead of .
Voicing (p vs. b, etc.) is not phonemic in Finnish, and due to this, voiced consonants may be devoiced. This includes:

Variation of voicing inside a long word shows the problems, even if short words are correctly pronounced. For example, bed, pet and bet may be distinct, but probably becomes propably or even propaply.
Devoicing of into , or confusing them when they appear in the same word. This is because in native Finnish pronunciation, 'b' is often pronounced identically to 'p', e.g. both baari (bar, pub) and paari (par) are pronounced paari. The mispronunciation bubi and pupi"pub" is found in Finnish.
Sometimes, and are both pronounced 'k', e.g. inkredient, where the "k" is a devoiced "g".
The English voiced sibilant becomes the unvoiced , e.g. roses becomes . When the spelling is 's', it is always pronounced 's'. When the spelling is 'z', attempts at voicing might be observed e.g. with haze vs. hays, but usually these are identical with the latter.
As in German, the letter 'z' is used to write in the Finnish alphabet. This may carry over to English, especially word-initially so that zealous might be pronounced . It may also be pronounced since Finnish speakers may confuse with the voiced affricate . This would make zealous and jealous indistinct.
Difficulty with both voiced (as in the ) and unvoiced (as in thin ) dental fricatives. Usually the pronunciation is not a fricative, but more like a dental or . For example, thick might be pronounced . While those speaking with other foreign accents may do so, Finnish speakers do not typically replace dental fricatives with alveolar fricatives ('s' or 'z').
Difficulty distinguishing between voiced and unvoiced ] postalveolars, e.g. pleasure is pronounced with either or . The affricate as in "jump" typically becomes unvoiced i.e. , or even alveolar ^*.
- The sound does not exist as a native phoneme in Finnish. While speakers typically have no trouble producing the sound, slip-ups may occur.
The Finnish set of vowels is extensive enough to contain the relevant English vowels, and results in little vowel confusion. However, of English pure vowels, Finnish recognizes only "tense" vowels , but does not recognize lax vowels . They are replaced by short "tense" vowels instead, thus there is no vowel quality contrast; in this respect, the Finnish accent works like Australian English. Thus, we have tick as and teak as [tiːk, unlike in native English, where the former is a lax vowel (Jeffers & Lehiste 1979:140).
In Finnish, there is no vowel reduction as there is in English. As a result, Finnish speakers put in non-reduced vowels in place of vowels that are typically reduced in native pronunciations. Examples:

There is no schwa in Finnish; typically the rounded cardinal vowel replaces it. The and may be transformed into Ö or Ä . The word "baker" might be instead of , or "bird" may become .
Similarly, when there is an 'a' in spelling but a schwa in pronunciation, it is pronounced . The word "American" might be pronounced instead of .
The back vowels and (the u in 'cut' and the ar in 'cart') both merge into become the latter. They do contrast in length, so that "cut" is pronounced while "cart" remains as is. In contrast, remains as is, as it is a different phoneme from the in Finnish.
- Vowel length allophony is reproduced reliably, as the difference is phonemic in Finnish.
Finnish speakers never use the glottal stop as an allophone of like many English speakers do.
The English "tense" and "lax" (i.e. vowels contrasted by quantity and advanced tongue root) are completely replaced with their respective quantity-contrasted vowels (short and long vowels), which differ only by length, not by quality like in English. E.g. "bit" is pronounced by native speakers but Finns typically say . Thus, the only difference between "bit" and "beat" is vowel length
Aspiration does not exist in Finnish and it is not heard, i.e. "top" and "stop" are differentiated only by the 's'. However, sometimes Finnish-speakers pronounce a geminate instead in the medial position. (Finnish phonetics do not allow initial or final geminates.) Geminates might also be used incorrectly, due the "generic gemination", e.g. Inttöönet instead of .
The English approximant R is usually pronounced rolled as in Finnish, or imitated in some innovative, but incorrect way, e.g. using a creaky voice instead of actual rhoticity. However, many speakers are able to produce the English R.
The sounds V and W are both reduced to the Finnish approximant V , which is something between them, i.e. a without any fricative quality.
Due to Finnish always stressing the first syllable, English words accented on other syllables are often stressed incorrectly, e.g. "vocabulary" is instead of the correct . This can produce confusion, especially with words like “record” where stress is contrastive.
In Finnish, there is only one pronoun for "he" and "she," and needing to remember to distinguish the two can cause slip-ups. There are also no articles like "the" and "a" and Finns may have difficulty remembering to include them when speaking.

French

The dental fricatives and are not phonemes in French:
- Speakers of Quebecois French often pronounce them as respectively.
- Speakers of Continental French tend to pronounce them as and .
- Some speakers may overcorrect and pronounce or as .
- Speakers may also pronounce them as and .
as in child, is often pronounced as as in Charlotte
as in jam, is often pronounced as as in Jacques
Speakers may drop , and/or they may overcorrect themselves and insert in front of words beginning with vowels.
The English rhotic may be replaced by the French equivalent; while a uvular fricative rhotic is typical for speakers in Paris, ^*," target="_blank" >and |[ are other possibilities depending on the speaker's dialect.
Unstressed vowels may not be reduced to schwa () in positions where they would be reduced in English, e.g., "Station" may be pronounced and "Stella" .
Stress timing may follow the French pattern, which places equal stress on each syllable (this may sound like stressing the final syllable to English speakers), or syllables may be unequally stressed but in a pattern that sharply conflicts with the standard English stress pattern.
Certain vowels that do not exist in French may be pronounced as their nearest French equivalents, including as , as , and as or . Speakers may also have difficulty with the English pattern of long and short vowels.

German

German does not have dental fricatives ( and ). Speakers may pronounce them as labiodental fricatives, and , or and ; the former is more likely with younger speakers.
is sometimes pronounced as or vice versa. Standard German lacks initial ; some dialects, however, such as Bavarian, have only and lack as a phoneme entirely.
Word-final voiced stops may be devoiced to sound like . This is usually the hardest to overcome.
as in jam, may be devoiced (especially by speakers from southern Germany), being pronounced as .
Since German adverbs are identical in form to adjectives, Germans often drop the ending -ly from English adverbs.
Lack of distinction between and : thus "man" and "men" are pronounced as the latter.
Speakers may not velarize in coda-positions as native speakers do (Jeffers & Lehiste 1979:140).
The German rhotic consonant varies from region to region so speakers from different areas will pronounce the English differently:
- Many German dialects, including most varieties of standard German, have a uvular , , or as in French and Danish. This is pronounced at the back of the throat, and in the case of the latter two may seem harsh or grating to native speakers of English.
- German speakers may use an alveolar as in Italian or Scottish. The German metal rock band Rammstein can be heard to use this in their music.
- Several German dialects are non-rhotic, this may carry over into a speaker's English pronunciation. In English, however, non-rhotic accents are found in most of England, the Southern Hemisphere (Australia, New Zealand, South Africa, etc.), Malta, most Caribbean islands, Singapore, Malaysia, as well as in Eastern North America (especially in costal regions), so this is less noticeable. Many German speakers also have a rhotic accent.
- Some German dialects use the same as native English speakers.
German speakers may pronounce as , and as . This is likely due to spelling pronunciation. However, Bavarian lacks , and Bavarians usually learn more easily than .
Because words with "qu" represents /kv/ in German spelling, speakers may mistakenly think that English is the same. Words like "quick" then sound like "kvick".
Inappropriate use of "he/she" pronouns for animals and inanimate objects. (German nouns have grammatical gender, so a foot is a "he", a child an "it", and a newspaper a "she".)
Difficulty in correct stressing of syllables; in most German words, the first syllable is stressed. Exceptions include words beginning with be-, ge-, er-, ver-, zer-, ent-, (and a few other prefixes); compound adverbs with her, hin, da, or wo; and loanwords to German from other languages. The stress normally falls on the second syllable in the former two examples.

Greek

Usually rolled
as in ship, is often pronounced as .
as in child, is often pronounced as .
as in jam, is often pronounced as .
is sometimes pronounced (like the Spanish: "José" or German: "Bach")
Voicless stops () are unaspirated.

Hebrew

The most common Hebrew dialect has only 5 vowels (though some distinguish between an additional four, normally pronounced the same) and generally does not use diphthongs (except for foreign borrowings); Hebrew speakers may therefore mispronounce some of the English vowels.
as in "the" is often pronounced as d (less commonly as z).
as in "think" is often pronounced as t, s or f.
Native English speakers typically velarize in words like "mill" . Hebrew speakers don't do this and their is always "soft" or "clear."
Rhotic accent. Historically, the Hebrew was an alveolar flap as in Spanish. Influence from European languages has led to Hebrews of European descent using a uvular , , or as in French or German.

Hungarian

Hungarian speakers are likely to exhibit regressive voicing assimilation, even across word boundaries. So, for instance, "truck driver" sounds more like "trugg driver" and "red house" like "ret house."
Most speakers have an absence of the English tonal variation. Falling intonation is used, making questions sound like statements.
Speakers may not use a glottal stop as an allophone of where native speakers would.
Speakers may replace schwa () with or .
Pronouncing instead of since the latter is not a phoneme in Hungarian.
Speakers may not make or hear a distinction between and , thus men and man are homophones.
Speakers may confuse the vowels and , e.g. bought and boat.
The dental fricatives and may be replaced by and Ádám Nádasdy: Background to English Pronunciation (Phonetics, Phonology, Spelling) – for students of English at Hungarian teacher training institutions
Speakers typically pronounce as since the velar nasal is an allophone of before velar consonants.
Speakers may have no distinction between dark and light as in English.
Speakers tend to pronounce like a short ^*.

The Indian Subcontinent

Note: There are many different languages and language families in India such as Bengali, Gujarati, Hindi, Urdu, Kannada, Marathi, Punjabi, and Tamil. Because of dominance by the British Empire for so long (See History of India) English is a commonly spoken language in India, especially in administration and as a Lingua Franca. Attempts at describing an "Indian" accent will naturally oversimplify variation that appears from one Indic language to another.

Use of the present continuous/progressive ("-ing") rather than simple present: "He has a car" becomes "He is having a car."
The rhotics or may be used instead of English (The former sound exists in some American dialects, though).
The English alveolar stops and , are likely to be perceived by many native Indic and Dravidian language speakers as the retroflex consonants and respectively. may also be retroflex. This leads to the "hollow" pronunciation of English by many Asian Indians.
Dental fricatives may be pronounced as dental stops.
The use of the double-positive, "Yeah, right," which in colloquial native English is a flippant way of saying "No" but to an Indian speaker of English is merely a double affirmation of correctness
Possible incorrect aspiration. Many Indic languages have distinct aspirated and plain pairs of consonants in addition to voicing; speakers may have difficulty with correctly identifying or producing allophones of when they are unaspirated.
Confusing 'p' and 'f' (among North Indians).
Difficulty with and . is not a common phoneme in Indic languages and speakers may produce both as . , a labiodental approximant, is another possibility.
Use of "isn't it?" in place of all other tag questions: "He is tall, isn't it?"
Vocabulary variations: "stay" instead of "live" (in a particular place), "cabin" instead of "office", "sit with" for "have a meeting with".

Icelandic

Difficulty with and .
Icelandic doesn't have or and speakers may pronounce them as and respectively.
Devoicing of the voiced consonants in word-final position.
is trilled in Icelandic. Speakers may have difficulty with the English .

Italian

Pronunciation of past tenses such as flipped as two syllables instead of the native pronunciation "flipt".
Tendency to pronounce the English high lax vowels as (ex: "fill" and "feel", "put" "poot" are homophones), since Italian doesn't have these vowels.
Tendency to pronounce as ("singer" rhymes with "finger") because Italian is an allophone of before velar stops.
Tendency to pronounce as . In English, the second consonant would assimilate the voicelessness of 's' but in Italian the process is reversed and assimilates the voicing.
Tendency to pronounce as , or so that "hut" may sound like "hat" or "hot", and "club" is typically pronounced as
Tendency to pronounce (British English) as ("cot" => "caught"). This is not as significant in North America since the two sounds have merged in many places (see Cot-caught merger). However, while in North America the phoneme has and as allophones, in Italian it is always an , except when speakers deliberately try to have an American accent, where an is often used.
Italian does not have dental fricatives:
- Voiceless may be replaced with a dental .
- Voiced may become a dental .
Since and are typically pronounced as dental stops anyway, words like "there" and "dare" can become homophones.
pronounced as . This and the above three tendencies make many Italians unable to pronounce death and that differently.
Tendency to pronounce as unaspirated stops.
Schwa does not exist in Italian and speakers may pronounce an 'i' or 'a' instead.
Due to the high prevalence of words ending in vowels in Italian, speakers may epenthesize vowels at the end of English words that end in consonants. This has led to the stereotype of Italian speakers adding an –a to every word ending in a consonant. In reality, this feature is more reminiscent of an occasional slip-up.
Tendency to pronounce words as they are spelled so that "aren't" is more like , "tomb" as and "chocolate" as .
In the diphthongs the second vowel is tense and lasts almost as long as the first. Sometimes they are pronounced as two syllables so that "toy" is and "cow" is
English 'o' is a diphthong . Italian speakers often pronounce it as .
Most Italian accents are rhotic, wherein 'r' is either a trill or a flap and speakers may have difficulty learning the English .
Italian doesn't have and speakers may not pronounce it.
Tendency to make double consonants geminate, e.g. the two ps in "happy" become a longer 'p' sound: This happens because geminate consonants are distinctive in Italian and, as mentioned above, Italian speakers tend to use spelling as a cue to pronunciation.
In Italian, the letter 'z' is used for the alveolar affricates and . While rare, it is possible for speakers to mistakenly carry this over to English pronunciation instead of .

Japanese

Tendency to confuse and , since they are indistinct in Japanese. Both are often pronounced as .
In Japanese, there is no . However, before , becomes and this sounds close to an . Speakers may have difficulty perceiving the difference between and .
Other than in front of Japanese speakers may produce instead of .
Japanese also lacks . Typically, and replace it (i.e. "video" becomes "bideo" or "wideo"). It is typically the former since is uncommon in Japanese.
Japanese has a largely consonant-vowel syllable structure (although there are some clusters that include prenasalized stops) and so both consonant clusters and words ending in consonants may have epenthetic vowels (usually , although it can be " after or ). Between voiceless consonants, these epenthetic vowels (and already existing vowels as well) may be voiceless.
May have trouble distinguishing 'm' and 'n' before bilabial consonants (b, m, & p)
Articles such as "the" and "a" are often dropped or used inappropriately.
Difficulty remembering to distinguish between singular and plural nouns.
There are a number of phonological processes that Japanese speakers may transfer to English:
- Alveolars become postalveolars in front of and .
  - becomes . (Some well-educated speakers hypercorrect their speech and make most instances of into .)
  - and become and .
  - The epenthetical vowel (mentioned above) after syllables ending in and is usually .
  - becomes like the German .
- Alveolar stops become affricated in front of .
  - and become and or ..
Possibly due to the inappropriate use of the gemination mark or sokuon when transliterating, speakers may occasionally geminate consonants.
Difficulty pronouncing sound except before 'a'.
Usage of pitch accent, as is normal in Japanese, rather than the standard English stress accent.
Dental fricatives are absent in Japanese and may be pronounced as alveolar fricatives or dental stops.
Often misrepresent schwa as a short .
Difficulty pronouncing the lax vowels , since these sounds do not exist in Japanese.
May reverse "yes" and "no" for a negative question. This is because the Japanese words for "yes" and "no" literally translate into "That's right" and "That's wrong" respectively. ("Didn't you bring your book?" "Yes, I didn't" meaning "You're right, I didn't.")

Korean

Difficulty distinguishing and since they are allophones of the same Korean letter. The phoneme may be pronounced in both the syllable-final and syllable-initial positions. Intervocalic may be pronounced . Most Korean speakers will speak with a rhotic accent.
Devoicing of initial (and sometimes final) consonants, since Korean does not have voiced initials or voiced finals.
The labiodental fricative sounds do not exist in Korean, so becomes and becomes .
Postalveolars do not exist as phonemes in Korean so the names 'Jack' and 'Zack' are likely to sound exactly the same.
Dental fricatives do not exist in Korean:
- may become or .
- may become or .
Tendency to epenthisize vowels on words ending with consonants. This is less likely with voiceless stops and sonorants. Usually the vowel is with syllables ending in fricatives. .
Tendency to break up consonant clusters with . Consonants may also be deleted in clusters or at the end of words
Plural markers are often omitted, because in Korean the plural markers are often omitted as long as context permits.
Korean has most of the vowels in English (Except for ) and ). However, in Korean, and are allophones of each other and Korean speakers may have trouble with accurately producing these sounds in English.
Speakers tend to ignore vowel reduction processes and pronounce words as they are spelled.
May reverse "yes" and "no" for a negative question. This is because the Korean words for "yes" and "no" are similar to "That's right" and "That's wrong" respectively. ("Didn't you bring your book?" "Yes, I didn't" meaning "You're right, I didn't.")

Latvian

Difficulty with /w/ and /v/.
Difficulty with articles such as “a” and “the”.

Liberia

All educated Liberians speak English, but for the native Liberians, their first languages are their native language, like Kru, Mandinka, and Fula.

/r/ is trilled and most accents are non-rhotic (also used by Afro-Americans and other native-born blacks).

Malay and Indonesian

Difficulty distinguishing between tense and lax vowels. Malay and Indonesian have only 6 vowel sounds, with no long-short distinction.
Dental fricatives are pronounced as stops.
Speakers may trill their r’s when speaking.

Mandarin (Chinese)

Mandarin phonology permits only , and in postvocalic positions. Therefore, speakers may have difficulty in the many words that have postvocalic consonants, often deleting them or following them with a vowel sound.
Trouble with dental fricatives and . They are typically pronounced as alveolar fricatives.
and are often confused since appears only as an allophone of in certain dialects of Mandarin.
Word-final devoicing of voiced consonants.
While both English and Mandarin distinguish between plain and aspirated stops rather than voicing, several allophones of English aspirated sounds are unaspirated and several allophones of plain stops are voiced. In Mandarin, this allophonic variation doesn't exist and Mandarin speakers may aspirate or devoice when native English speakers would not.
Tonalization of English syllables, as Mandarin Chinese is a tonal language; Usually a combination of high and falling for stressed syllables, as well as neutral and low tones for unstressed syllables.
Variation in English syllable length and timing from standard native English, due to the timing of Mandarin (see syllable timing)
Speakers may have difficulty with pronoun gender: (he, she, and it) since spoken Chinese does not make gender distinction in pronouns (although these words have different hànzì).
Confusion of and . Mandarin Chinese has no final sound or initial Pinyin as r) that sounds similar to [r" target="_blank" >^*. Most Chinese speakers speak a non-rhotic or partly non-rhotic accent.

See also: Chinglish

Maori

is a tap, like with Spanish. Typically speech is non-rhotic
Aspirated consonants are unaspirated, sounding like 'b' 'd' and 'g.'

See also: New Zealand English

Mozambique

Although most black Mozambicans speak Portuguese, their first languages are native Bantu languages, like Makonde, Tsonga, Shona, and Makua. So the English accent of educated black Mozambicans is affected by these languages.

Dental fricatives are pronounced as corresponding plosives
/r/ is trilled and most accents are non-rhotic.

Nigeria

Note:Many Nigerians are educated entirely in English, but their first language, spoken at home, is a native Nigerian language, such as Yoruba, Hausa, Igbo or Ibibio.

and tend to be replaced by and .
Non-rhotic. The actual rhotic consonant varies but may be an alveolar flap, trill, or just like the English . When it is the latter of the three, the accent would sound more like native British were it not for the vowel inventories of Nigerian languages.
Several vowels present in English are absent from Nigerian languages. There is usually variation from region to region on what is used
- "Bird" ( in RP) is pronounced variably as or . In the west of the country, is more common , while is more common in the East.
- The vowel is replaced by or , so "cut" is pronounced or , the latter being Eastern.
- "O," which is a diphthong in most dialects of English ( in RP, in North American), is generally pronounced as ,
- The diphthong is typically pronounced as .
Some native Yoruba speakers tend to put at the beginning of words that begin with a vowel.
Some native Hausa speakers pronounce as .
Some speakers replace with , so "yet" is pronounced like "jet."
Speakers may replace with in certain contexts, pronouncing "watch" as "wash."
There is a syllabic rhythm to speech, and unstressed syllables are often pronounced distinctly.
Words from Nigerian languages are often injected into speech.
Questions are sometimes constructed as a statement followed by "sha" and a rising tone. Another way of constructing a question is to use "shebi" or "abi" as a marker at the beginning of the sentence.
The meanings of "no" and "yes" are reversed for negative questions relative to standard English ("Yes, I have no bananas"). People often say "I'm coming" as they go away, meaning they'll be back soon.

Persian

Persian has no dental fricatives, and speakers tend to substitute stops in their place.
Since Persian has no , is commonly used instead. It may also be realized as a labiodental approximant , a sound somewhat in between and .
In Persian, /r/ is an alveolar trill in the initial position, a tap intervocalically and a voiceless trill word-finally (Ladefoged & Maddieson 1996:216). This may carry over when attempting to produce the English rhotic .
Speakers often lengthen the final vowel in long words while stressing the first syllable.
Persian does not have a vowel reduction pattern like English does, and as a result speakers may have difficulty producing schwa.
Consonant clusters that Persian speakers find difficult may be separated by epenthetical vowels or simplified through deletion. The epenthetical vowel may be at the beginning of the word in cases of words like "stop" to make it sound more like "e-stop" and deletion is commonly at the end of words making "five" sound like "fie."
Final -ng may become -nk as in "studyink economics".

Polish

Polish has no dental fricatives. Apical stops or labiodental are typically substituted. In the past, alveolar fricatives were known to be used. This results in identical pronunciations of minimal pairs such as thought/fought and that/vat.
Voiced consonants may become voiceless in word-final positions or may assimilate devoicing when next to voiceless consonants. This results in identical pronunciations of minimal pairs such as pig/pick or bag/back.
Trouble differentiating between similar vowels:
- and sound similar, although the latter could be perceived as , which is a phoneme in Polish.
- and may be the former.
- and are not phonemes in Polish. Speakers may produce spelling pronunciation (i.e. and respectively) or approximate the vowels some other way.
- and are pronounced as the former. E.g. "man" and "men" are pronounced .
A few commonly used false friends, most prominently "actually" with intended meaning of "currently".
Problems with articles, since Polish does not have them.
Inappropriate use of "he/she" pronouns for animals and inanimate objects.

Portuguese

Portuguese has no dental fricatives. Speakers may pronounce them as dental stops.
- Voiceless may be replaced with a dental or, more commonly, with
- Voiced may become a dental .
Since and are typically pronounced as dental stops anyway, words like "there" and "dare" can become homophones.
pronounced as . This and the above three tendencies, make many Portuguese speakers unable to pronounce death and that differently.
Tendency to pronounce as unaspirated stops.
Common mistakes due to Portuguese cognates with different meanings ("false-friends"):
- "Deception" used when the intended word was "disappointment."
- "Actual" used when the intention was "current."
- "To pretend" used when the intention was "to intend."
Portuguese has no and it may be substituted by . However, it could be perceived as , which is a phoneme in European Portuguese.
Speakers may make spelling pronunciation mistakes:
- Occasionally pronounce 'oo' in "door" or "Moore" as
- May pronounce "apple" as
Typically speech is non-rhotic (except Brazilian and some other dialects), having no in the coda. Depending on where the speaker is from, it may be an alveolar or a uvular .
Postalveolar fricatives usually substitute postalveolar affricates.
is typically dropped.
In Brazil, Depending on the speaker's region, alveolar stops and may be palatalized before so that they sound more like and

Romanian

Romanian does not have dental fricatives:
- Voiceless may be replaced with a dental .
- Voiced may become a dental .
Since and are typically pronounced as dental stops anyway, words like "there" and "dare" can become homophones.
pronounced as . This and the above three tendencies, make many Romanians pronounce death and that as homophones.
Tendency to pronounce the English high lax vowels as (ex: "fill" and "feel", "put" "poot", "live", "leave" are homophones), since Romanian doesn't have these vowels.
Tendency to pronounce as ("singer" rhymes with "finger") because Romanian is an allophone of before velar stops.
Tendency to replace the English retroflex r with the Romanian trilled /r/.
Common mistakes due to Romanian cognates with different meanings ("false-friends"):
- "Eventually" used when the intended word was "probably."
- "Actual" used when the intention was "current."
Some speakers may have difficulty in using prepositions. Romanian is an inflected language and use of prepositions is less extensive than in English. While English prepositions modify the meaning of verbs, in Romanian there are separate words for each meaning (e.g. "go in" = "a intra", "go out" = "a ieşi", "go away" = "a pleca", "go ahead" = "a continua".)
A significant number of Latinate words have identical or very similar spelling in Romanian and English, making pronunciation confusing.

Russian

There is no in Russian and speakers typically substitute and will have trouble perceiving the difference between the two.
There are no dental fricatives in Russian, so they may be replaced with alveolar fricatives or dental stops.
Alveolar consonants may be pronounced as dental.
Fortis plosives may be pronounced plain and Lenis plosives as fully voiced (except word finally; see below), whereas native speakers usually aspirate fortis stops and devoice lenis stops.
In Russian, every consonant is either soft or hard. If a consonant appears before , or it is likely to be palatalized. This is most noticeable with , , and as their place of articulation changes from dental or alveolar to postalveolar (which then affricates) or even palatal. Palatalization in cognates may also be a factor: dune, дюна .
Russian is an alveolar trill. This may carry over into English. Speakers able to pronounce English typically velarize it so it comes out as . may also be a velarized
Voiced consonants may be devoiced in word-final position, this is normally done in Russian.
may be pronounced as a velar fricative , which does not exist in most dialects of English.
Russian has only five or six vowels and speakers may have trouble with vowels not in their native inventories.
- and are pronounced as the former. E.g. “man” and “men” are pronounced .
- The diphthongs , , and sound with the consonant sound instead of the short . E.g. “high” sounds like , rather than .
Inappropriate use of “he”/“she” when referring to animals or inanimate objects.
Difficulty with articles such as “a” and “the”.

Spanish

Spanish lacks many vowels present in English, often causing the following changes:
- becomes , or even
- becomes
- becomes
- becomes the vowel that corresponds to the written letter.
- becomes
- (British) or (American) becomes
- In general the pronunciation of vowels will tend to follow the Spanish pronunciation of their written form.
- Most English diphthongs have equivalents in Spanish, although the lack of lax vowels means that, for example becomes .
Voiceless stops are unaspirated, making them sound like voiced stops to native speakers.
A few dialects of Castilian and American Spanish have both y and ll as distinct phonemes. Most Castilian and American dialects have lost the distinction (see Yeísmo) and speakers pronounce both as either , , or . Usually, speakers will substitute one of these for English , , and .
is pronounced as by many speakers, since the former does not exist in most Spanish dialects.
Intervocalic and final , , and , become the corresponding approximants , , and , as it is done in standard and Latin American Spanish. As a result of this, certain English phonemes are dealt with as allophones.
- Both and become or (depending on the speaker's region) unless in a word-initial position, where they are realized as .
- Speakers tend to pronounce as a voiced dental stop unless it is intervocalic and as when it is intervocalic (Jeffers & Lehiste 1979:139).
Spanish has a trilled and an alveolar tap, . Speakers may have the former word-initially and the latter intervocalically.
Most accents are rhotic. There are a few exceptions such as in Gibraltar and some dialects of Castilian.
European speakers may not have trouble with (see ceceo) but other speakers may pronounce it as a voiceless dental stop. Speakers may also pronounce it as .
is often devoiced to .
Often a word beginning with + consonant will obtain an epenthetical vowel (typically or ) to make "sprite" sound more like "esprite". In Spanish, /s/ must immediately precede or follow a vowel (some speakers may hypercorrect themselves and say "stir" for "ester").
Occasional mispronunciation of final as , e.g. "welcome" -> "welcon". This is how it is pronounced in the few Spanish words ending in "m" (most notably, "álbum" and "réquiem").
Pronunciation of final , especially before non-velar consonants, as (sing -> seen), because is an allophone of before velar stops.
is either pronounced as or dropped. The former is because the closest Spanish sound to in most dialects is ; the latter because orthographic is silent in Spanish. In some dialects, such as Colombian, has become and speakers usually have no trouble with English .
Syllable-final may sometimes turn into or get dropped altogether, as it is done in some Spanish dialects.

Swahili

Devoicing of consonants in word-final position
Some speakers have trouble with dental fricatives, pronouncing them as either dental stops or alveolar fricatives. This is not uniform.
Some speakers have difficulty distinguishing between and . Most dialects are non-rhotic and the rhotic consonant for Swahili speakers is an alveolar flap
Stress is often placed on the penultimate syllable: newspaper is newspaper.

Swedish

Sing-songy intonation. Swedes often speak English with a melodic intonation, ending sentences on an up-note, much parodied (The Swedish Chef from The Muppet Show is a well known example.)
Difficulty pronouncing dental fricatives. Usually a dental stop or labiodental fricative is substituted.
Swedish lacks a , it is often pronounced .
Swedish lacks a , it is often pronounced .
is often pronounced as . This could be caused by a mix-up with Swedish orthography, where ch often has a sound more like English sh.
as in jump is sometimes pronounced as (pleasure) or even (yellow), (the latter could be caused by mix-up with Swedish orthography, where j has this sound).
Some difficulty in verb conjugations (e.g. "they is"). Swedish verbs do not inflect for person.
Different Swedish dialects have different realizations of /r/. They range from alveolar trills and flaps to uvular sounds and even retroflex fricatives.
The sound when written as ti (as in nation) may be pronounced as . This could be due to problems with orthography.
Diphthongs: may become , and would become . The former may not be significant and the latter may be due to problems with orthography.
Words written with rs, rt, rd, rn and rl often turn to retroflex variants of the consonant following r.
Swedish does not have dental fricatives. Most commonly, becomes ("though" sounds like "dough") and becomes or ("think" sounds like "sink" or "fink").

See also: Swenglish

Swiss German

Voiced consonants may be devoiced in all positions, especially word-finally — lenis and fortis plosives may differ in length and aspiration, yet Swiss German dialects do not have any fully voiced plosives. This is not too different from English
Some dialects have instead of aspirated , hence some speakers use this sound.
often pronounced like or ; Swiss dialects don't have dental fricatives
Dialects that have an phoneme (e.g. Basel, Zurich) usually use this to represent English ; others (e.g. Graubuenden) however use , which can then create homophones.
Older speakers may use instead of ; other speakers generally use a variant of

Tagálog/Filipino

Tagálog has only five vowels, so the many vowel sounds in English may be mapped to the nearest-sounding existing vowel. However, a number of English vowels appear as allophones or in unstressed positions and this may facilitate a speaker's ability to learn the vowels.
Certain English phonemes missing in Tagálog may be replaced by similar sounding phonemes in Tagálog:
- Labiodental fricatives may become bilabial plosives.
- Dental fricatives may become dental plosives.
- may become .
- may become .
- may become .
- may become .
- Tagálog 'r' is an alveolar flap. Speakers may have difficulty pronouncing the English .
Tendency to epenthesize vowels when dealing with consonant clusters. (e.g., sport may become is-port).
Speakers may mix up "he," "she," and "it." Tagálog does not distinguish gender in pronouns.
Rhotic speech, so 'r' is pronounced even in the coda of a syllable. (This happens in native English pronunciations of Barbados, Ireland, Scotland, most of Canada and the United States, and parts of England as well.)

Thai

may be trilled instead of the approximant
non-rhotic speech
at the end of a word may be pronounced as or
, , and can be indistinguishable and pronounced as
may be pronounced as or
may be devoiced to (devoiced)
may be devoiced to (devoiced)
may be pronounced as word-initially and word-finally.
Thai has no dental fricatives and usually substitutes apical stops for them.
English syllables may be tonalised since Thai is a tonal language.

See also Tinglish

Turkish

Turkish lacks any grammatical gender, so a female person may be accidentally referred to as "he" and a male as "she".
Final obstruent devoicing.
Dental fricatives become dental stops.
becomes and/or .
Speakers may trill their r’s when speaking.
clusters including liquids (those including /l/ and /r/) may have an epenthetical schwa.

Ukrainian

The Ukrainian voiced or velar may be used in place of the English .
and may both be pronounced as .
Speakers may use a trilled instead of the English approximant.
Lack of distinction between and .
Lack of distinction between and .
may be dental instead of alveolar; these sounds may also be substituted for the English dental fricatives .

Vietnamese

Note: There are two main dialects in Vietnamese, a northern one centered around Hanoi and a southern one centered around Ho Chi Minh City.

Speakers may epenthesize a glottal stop before words beginning with vowels. For northern speakers, this extends to words beginning with /w/.
Extensive devoicing of consonants, especially word-finally.
Complex cluster simplification through either epenthesizing vowels or through consonant deletion.
Coda obstruents may be deleted.
Vietnamese has implosive sounds and English and may be pronounced as and respectively in the onset of a syllable.
Vietnamese does not have dental fricatives and these may be pronounced as dental or alveolar plosives.
English syllables may be tonalized since Vietnamese is a tonal language.

References

External links

A site collecting recordings of people from different areas reading the same paragraph (most recordings also have an IPA transcription)
International Dialects of English Archive

Pages containing IPA | Forms of English | English phonology

Gourt Home::Gourt :: Non-native pronunciations of English