The following tables show the International Phonetic Alphabet (IPA) and the English pronunciation (enPR) or American Heritage Dictionary (AHD) symbols that are used to represent the various sounds of the English language. The sounds of Received Pronunciation (RP, UK), General American pronunciation (GenAm, US), Canadian English (CanE), Australian English (AuE), and New Zealand English (NZE) are shown.
For vowels in other dialects, see Wikipedia's IPA chart for English. An image of an old version of these tables is available.
For a fuller list of dialects, see International Phonetic Alphabet chart for English dialects.
Non-rhotic dialects (RP, Australia, New Zealand) do not pronounce what was historically a syllable-final r; instead they have a schwa (/ə/), centering diphthong (ending in /ə̯/), or a long vowel (ending in ː). Rhotic dialects (General American and Canadian) pronounce the syllable-final r. Also, they make no vowel length distinction, so none of their vowels end in the length symbol ː.
This vowel table below lists the standard phonemic vowel notations in each accent and contains both monophthongs and diphthongs. Variations of notation within the same accent are also listed.
enPR / AHD[1] | IPA | examples | ||||
---|---|---|---|---|---|---|
RP | GenAm | CanE | AuE | NZE | ||
ä | ɑː | ɑ | ɒ, ɑ | ɐː, aː | father, palm | |
är | ɑː | ɑɹ | ɐː, aː | arm, bard, starry, start | ||
ă | æ, a | æ | ɛ | bad, cat, ran, trap[2][3] | ||
ăr | æɹ, aɹ | æɹ, ɛɹ | ɛɹ | æɹ | ɛɹ | carry, marry, paragraph[4] |
ā | eɪ | æɪ | day, pain, hey, weight, face | |||
âr | ɛə | eɹ, ɛɹ | ɛɹ | eː | iə, eə | hair, there, fairy, vary, Mary, square[5][6][7][4] |
ĕ | ɛ | e | bed, egg, meadow, dress[8] | |||
ĕr | ɛɹ | eɹ | very, error, merry[4] | |||
ē | iː | i | iː | ease, see, siege, ceiling, fleece | ||
îr | ɪə, ɪː | ɪɹ | iɹ | ɪə | iə, iːə | near, here, serious[6][7] |
(ē) | ɪ, i, iː | i | iː | city, happy, everyday, mania, geography[9] | ||
ĭ | ɪ | ə | sit, city, bit, will, kit | |||
(ĭ) | ɪ | ə | roses, spotted, secure | |||
ĭr | ɪɹ | əɹ | mirror, Sirius | |||
ī | aɪ | aɪ (ʌɪ) | ɑɪ | aɪ | my, rice, pie, hi, Mayan, price[10] | |
ŏ | ɒ | ɑ | ɒ, ɑ | ɔ, ɒ | not, wasp, cot, lot | |
ŏr | ɒɹ | ɑɹ (rarely oɹ) | oɹ | ɔɹ, ɒɹ | borrow, sorry, sorrow, tomorrow (sometimes morrow)[11] | |
ŏr | ɒɹ | oɹ (regionally ɑɹ) | oɹ | ɔɹ, ɒɹ | horror, forest, orange, quarrel, warrior[11][12][13] | |
ôr | ɔːɹ | oɹ | oːɹ | forum, glory | ||
ôr | ɔː | oɹ | oː | horse, north,[6] hoarse, force[13] | ||
ô | ɔː | ɔ (with the cot-caught merger: ɑ) | ɒ, ɑ | oː | law, caught, thought[14] | |
ō | əʊ | oʊ | əʉ, ɐʉ | ɐʉ | no, go, hope, know, toe, goat | |
oi | ɔɪ | oɪ | boy, noise, choice | |||
o͝o, ŏŏ | ʊ | put, foot, wolf | ||||
o͝or, ŏŏr | ʊə, ɔː | ʊɹ | ʊə | ʉə, ʉːə, oː | tour, tourism, stoor, cure[6] | |
o͞o, ōō | uː | u | ʉː | lose, soon, through, goose | ||
ou | aʊ | æɔ | æʊ | house, now, tower, mouth | ||
ŭ | ʌ | ɐ, a | run, enough, up, other, strut[15] | |||
ûr | ɜː | ɜɹ | ɜː | øː | fur, blurry, bird, swerve, nurse[16][17] | |
ə | ə | Rosa's, about, oppose, comma | ||||
ər | ə | əɹ | ə | ə, ɐ, a | winner, enter, error, doctor, letter[16][17] |
Found primarily in some English dictionaries' transcription of the original (foreign) pronunciations of foreign words, especially French or German.
enPR / AHD[1] | IPA | examples |
---|---|---|
œ | œ, ø | oeuvre, Königsberg |
ü | y | Debussy, Württemberg (also used in some dialects, e.g. the MLE pronunciation of wound) |
In order to allow Module:syllables to count syllables, the disyllabic sequence /iə/ must be transcribed with a period to mark the syllable break – /i.ə/ – so that it will not be confused with the New Zealand diphthong /iə/.
Syllable-final /ɹ/ is sometimes replaced with /ɚ/: /ðɛɚ/ instead of /ðɛɹ/. In order to keep Module:syllables from counting /ɚ/ as a syllable, add the non-syllabic diacritic: /ðɛɚ̯/.
Some speakers do not contrast unstressed /ɪ/ and /ə/, or the two sounds may be in free variation. Some sources use the symbol ⟨ɨ⟩ or ⟨ᵻ⟩ to indicate the vowel which results from the merger of, or which may be pronounced as either of, these sounds.[1][2]
To be added or sorted into the table above: bath, cloth.
IPA | enPR / AHD[3] | examples |
---|---|---|
b | b | but, web, rubble |
t͡ʃ | ch | chat, teach, nature |
d | d | dot, idea, nod |
f | f | fan, left, enough, photo |
ɡ | g | get, bag |
h | h | ham |
ʍ (hw)[4] | hw | which |
d͡ʒ | j | joy, agile, age |
k | k | cat, tack |
x | ᴋʜ[5] | loch (Scottish English) |
l | l | left |
l̩ (əl)[6] | l | little |
m | m | man, animal, him |
m̩ (əm)[6] | m | spasm, prism |
n | n | note, ant, pan |
n̩ (ən)[6] | n | hidden |
ŋ | ng | singer, ring |
p | p | pen, spin, top, apple |
ɹ[7] | r | run, very |
s | s | set, list, ice |
ʃ | sh | ash, sure, ration |
t | t | ton, butt |
θ | th | thin, nothing, moth |
ð | th | this, father, clothe |
v | v | voice, navel |
w | w | wet |
j | y | yes |
z | z | zoo, quiz, rose |
ʒ | zh | vision, treasure |
The so-called voiceless and voiced obstruents are more properly fortis and lenis. Each member of a fortis–lenis pair is distinguished from the other by various articulatory and auditory features, but not consistently by voicing or lack of it.
In most dialects of English, the fortis (voiceless) stops and affricate /p t tʃ k/ are always voiceless, and are aspirated () at the beginning of a word and at the beginning of a stressed syllable: for example, RP today , chain and account . Vowels and sonorants immediately preceding syllable final fortis obstruents are usually pronounced shorter than before lenis obstruents, as in bet vs. bed and bent vs. bend. This phenomenon is known as pre-fortis clipping.
The lenis (voiced) stops and affricate /b d dʒ ɡ/ are always unaspirated. Lenis obstruents /b v ð d z dʒ ʒ ɡ/ are often devoiced at the beginning or end of words, but are fully voiced between voiced vowels and sonorants.
The fortis–lenis distinction is neutralized in a few cases.
Initial consonant clusters consisting of /s/ and a stop (as in spill, still, skill) are typically analyzed as having a fortis stop, which agrees with the spelling, but may be analyzed equally well as having a lenis stop (i.e., *sbill, *sdill, *sgill). The stop is both voiceless and unaspirated, and there is no additional phonetic feature that establishes it as either fortis or lenis.
In addition, some dialects have a sound change known as intervocalic alveolar flapping, in which /t d/ are both pronounced as an alveolar flap between vowels or liquids and when not at the beginning of a stressed syllable. Further, in American English, /nt/ between vowels may be pronounced as a nasalized alveolar flap, . The fortis stop /t/ loses its distinctive voicelessness, and essentially becomes lenis. Flapping causes latter and ladder to both be pronounced as , and causes winter to be pronounced as , similar to winner .
When two vowels occur next to each other (termed hiatus), speakers sometimes perceive the vowels to be separated by a sound similar to one of the semivowels /j w/. The identity of such "linking semivowels" is predictable based on the identity of the preceding vowel: a /j/-like sound may be perceived after vowels ending in a high front unrounded sound, such as /iː~i/, /ɔɪ~oɪ/, /aɪ~ʌɪ~ɑɪ/, /eɪ~æɪ/, whereas a /w/-like sound may be perceived after vowels ending in a high back (or central) rounded sound, such as /uː~u~ʉː/, /aʊ~æʊ/, /əʊ~oʊ~əʉ~ɐʉ/.
Even though some speakers hear semivowels in these contexts, there is evidence that such "linking" semivowels are not phonetically identical to the semivowel phonemes that can be found at the start of words (as in "yearn" /jɜː(r)n/ or "weevil" /wiːvəl/). For example, the phonetician John Wells discusses "I earn" versus "I yearn" and "two evils" versus "two weevils" as minimal pairs showing that there is usually no neutralization of the phonemic contrast between the sequences /ɑɪ.ɜː/ and /ɑɪ.jɜː/, or between /tuː.iː/ and /tuː.wiː/.[1][2] Therefore, such "linking semivowels" should not be included in phonemic transcriptions.
It is also inadvisable to include them in phonetic transcriptions, since a number of phoneticians have argued that what is heard as a semivowel is actually nothing more than the final portion of the preceding vowel or diphthong.[1][2] For example, assuming we transcribe yellow and ready as and , it is unnecessary and redundant to use transcriptions such as yellowing or readying as opposed to and , since there is likely no distinction on the phonetic level between the segments transcribed , and the supposedly following segments , . A 2014 phonetic study of American English found that there were significant acoustic differences between the pronunciation of two vowels separated by a phonemic glide, and sequences of a high vowel or diphthong followed by a vowel: the perceptual illusion of a glide in the latter case could be explained in terms of a diphthongal realization of the first vowel phoneme, rather than insertion of a glide after it.[3]
An alternative analysis of the English vowel system treats the glides /j w/ as an inherent part of diphthongs and "tense" vowels.[4] In this kind of analysis (which is not used on Wiktionary), words like yellow, yellowing and ready, readying might be transcribed as /ˈjɛləw/, /ˈjɛləwɪŋ/ and /ˈɹɛdɪj/, /ˈɹɛdɪjɪŋ/. Note that this analysis also does not involve glide insertion when these vowels are placed before other vowels.
A stress mark is placed before the syllable that is stressed in IPA and after it in enPR / AHD.
IPA | enPR (AHD) |
indicates |
---|---|---|
ˈ (ˈa) | ′ (a′) | primary stress, as in rapping /ˈɹæpɪŋ/ |
ˌ (ˌa) | ′ (a′) | secondary stress when before the primary stress; unstressed full vowel when after the primary stress, as in battlefield (phonetically , phonologically /ˈbætəlfiːld/) |
a.a | a-a | division between syllables |
̩ | syllabic consonant, as in ridden | |
ʔ | glottal stop, as in uh-oh /ˈʌʔoʊ/, | |
̃ (ã) | ɴ (aɴ) | nasalization, as in croissant /ˈkɹwæsɒ̃/ |
Note: The EnPR and print AHD marks are formatted slightly differently. Online, AHD writes both ', though they do not always represent the same phoneme.