Pronunciation respelling for English

From Infogalactic: the planetary knowledge core
Jump to: navigation, search

Pronunciation respelling is a notation used to convey the pronunciation of words, in a language, such as English, which does not have a phonemic orthography.

There are two basic types of pronunciation respelling:

  • "Phonemic" systems, as commonly found in American dictionaries, consistently use one symbol per English phoneme. These systems are conceptually equivalent to the International Phonetic Alphabet (IPA) commonly used in bilingual dictionaries and scholarly writings, but tend to use symbols based on English rather than Romance-language spelling conventions (e.g. ē for IPA /i/) and avoid non-alphabetic symbols (e.g. sh for IPA /ʃ/).
  • On the other hand, "non-phonemic"[1] or "newspaper"[2] systems, commonly used in newspapers and other non-technical writings, avoid diacritics and literally "respell" words making use of well-known English words and spelling conventions, even though the resulting system may not have a one-to-one mapping between symbols and sounds.

As an example, the last name of actor Jake Gyllenhaal, written /ˈdʒɪlənhɔːl/ in the IPA, might be written jĭl′·ən·hôl or JIL-ən-hawl in a phonemic system, and Jill-in-hall in a non-phonemic system.

Development and use

Pronunciation respelling systems for English have been developed primarily for use in dictionaries. They are used there because it is not possible to predict with certainty the sound of a written English word from its spelling or the spelling of a spoken English word from its sound. So readers looking up an unfamiliar word in a dictionary may find, on seeing the pronunciation respelling, that the word is in fact already known to them orally. By the same token, those who hear an unfamiliar spoken word may see several possible matches in a dictionary and must rely on the pronunciation respellings to find the correct match.[3]

Traditional respelling systems for English use only the 26 ordinary letters of the Latin alphabet with diacritics, and are meant to be easy for native readers to understand. English dictionaries have used various such respelling systems to convey phonemic representations of the spoken word since Samuel Johnson published his Dictionary of the English Language in 1755. Today, such systems remain in use in American dictionaries for native English speakers,[4] but they have been replaced by the International Phonetic Alphabet (IPA) in linguistics references and many bilingual dictionaries published outside the United States.[5]

The pronunciation which dictionaries refer to is some chosen "normal" one, thereby excluding other regional accents or dialect pronunciation. In England this standard is normally the so-called Received Pronunciation, based upon the educated speech of southern England. The standard for American English is known as General American (GA).

Sophisticated phonetic systems have been developed, such as James Murray's scheme for the original Oxford English Dictionary, and the IPA, which replaced it in later editions and has been adopted by many British and international dictionaries. The IPA system is not a respelling system, because it uses symbols not in the English alphabet, such as ð and θ. Most current British dictionaries[6] use IPA for this purpose.

Traditional respelling systems

The following chart matches the IPA symbols used to represent the sounds of the English language with the phonetic symbols used in several dictionaries, a majority of which transcribe American English.

These works adhere (for the most part) to the one-symbol-per-sound principle. Other works not included here, such as Webster's New Twentieth Century Dictionary of the English Language (unabridged, 2nd ed.), do not, and thus have several different symbols for the same sound (partly to allow for different phonemic mergers and splits).

The full titles of abbreviated column headings in the following table are viewable in interactive media (as opposed to hard copy), using the pointer. To see the full titles, hover over the abbreviations. On touchscreens, a long press and cancel may show them.

č ᴄʜ ch c͡h ch ch ch ch ch ch ch ch ch ch ch ch ch ch, tch ch ch church
h h h h h h h h h h h h h h h h h hh h h h h hat
hw hw hw (h)w hw hw hw hw hw hw   hw (h)w w hw hw wh which
ǯ j j j j j j j j j j j j j j jh j j j j judge
x x x ᴋʜ KH ᴋ͡ʜ kh H k (χ) kh hh ᴋʜ kh kh kh loch (Scottish)
Buch (German)
ɬ ɬ ł hl hl llan (Welsh)
ɡ9 g g g g g g g g g g g g g g g g g g g g g, gh game
ŋ ŋ ŋ ɴɢ ng n͡g ng ng ŋ ng ng ng ŋ ng ng ng ng ng ng ng
(ng-g, nk)
ng ng thing
ʃ ʃ š ꜱʜ sh s͡h sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh ship
θ θ θ ᴛʜ th t͡h th th th th th th th th th th th th th th th th thin
ð ð ð ᴛʜ th th̸ th th th t̷h TH th: th dh th dh TH dh th dh dh dh this
j j y y y y y y y y y y y y y y y y y y y y yes
ʒ ʒ ž ᴢʜ zh z͡h zh zh zh zh zh zh zh zh zh zh zh zh zh zh zh zh vision
The following letters have the same values in all systems listed: b, d, f, k, l, m, n, p, r1, s, t, v, w, z.
æ æ æ a ă a a a a a a a a ă a a a ae a a (arr) ă a (arr) cat
e e(y) ā ā ā ay ay ā ay ā ay ā ā ay ā ay, ae ey ey ay ā ay day
ɛər ɛr εr e(ə)r âr âr air air - - er air ar ār air ār air eh r air air ār air hair
ɑː ɑ a ä ä ä ah aa ä aw, o ä ah ä, ȧ ah aa ä ah aa ah aa â ah father
ɑr ɑr ar är är är ahr aar är - är ahr är a͡r aar är ar aa r ahr ar âr ar arm
ɛ ɛ ε e ĕ e eh e e ɛ e e e ě e e e eh e e (err) ĕ e (err) let
i i(y) ē ē ē ee ee ē ē ē ee ē ē ee ē ee iy ee ee ē ee see
ɪər ɪr ιr i(ə)r îr ēr ihr eer ir ir ēr eer īr ihr ih r eer eer ēr eer here
ɪ ɪ ɪ i ĭ i ih i i i i i i ǐ i i i ih i i (irr) ĭ i (irr) pit
ay ī ī ī y ī ī y ī igh ī ī ī ī eye, ie, ye ay ahy y ī y, eye6 by
ɒ ɑ a ä ŏ o o o ä o o ah ä ǒ o o o aa o o (orr) ŏ o (orr) pot
o o(w) ō ō ō oh ō ō ō ō oh ō ō ō ō oh, oe ow oh oh ō oh no
ɔː ɔ ɔ ô ô ô aw aw ô aw, o ô aw ȯ aw aw ö aw ao aw aw ô aw caught
ɔr ɔr ɔr ôr ôr ôr ȯr or awr or ôr awr, or north
ɔər or o(w)r awr, ohr ōr ohr, or force
ɔɪ ɔɪ ɔy oi oi oi oy oy oi oy oi oi ȯi oi oy oi oi oy oi oy oy oy noise
ʊ o͝o o͝o o͝o u o͝o oo u oo o͝o o͝o ŭ2 u uh oo uu ŏŏ uu took
ʊər ᴜr ᴜr o͝or o͝or o͝or ur oor u̇r u̇r oor oor oor uh r oor oor oor oor tour
u u(w) o͞o o͞o o͞o oo oo o͞o ū ü oo: ü o͞o o͞o oo2 oo uw oo oo ōō oo soon
aᴜ aw ou ou ou ow ow ou ow ou ow au̇ ow ow ow ou aw ou ow ow ow out
ʌ ʌ ʌ ə ŭ u uh u u ʋ u UH ə ǔ u u uh ah uh u (urr) ŭ u (urr) cut
ɜr, ɝː ɜr ər ər ûr ûr ur ur ʉr er ėr ER ər e͡r ər ûr ur er ur ur ûr ur word
ə ə ə ə ə ə uh ə ə e ə uh ə a, e, i, o, u ə ə uh ah uh uh ə ə about
ɨ ɪ, ə ə ə, i ə ə uh ə ə e ə uh ə a, e, i, o, u ə ə uh ah uh i ı i, ə7 rabbit
ər ɚ ər ər ər ər uhr ər ər er ər er ər er ər ər ur er er uhr ər ər butter
juː ju yu yo͞o yo͞o yo͞o yoo yoo yo͞o yoo: ū yo͞o ū yoo y uw yoo yoo ū ew view
a ȧ ȧ   ȧ ami (French)
øː, œ œ ɔ̈ œ œ ɶ œ œ̄, œ ɶ oe ö feu (French)
schön (German)
yː, ʏ y ü ʏ ü ʏ ʏ u͞e, ue ʏ ue ü tu (French)
über (German)
ɔ̃ õ ɔ̨ ôɴ ôɴ ôɴ ôɴ ōⁿ awɴ o(ng) ôⁿ bon (French)3
ˈa ˋa á ˈa aʹ aʹ ˈa a. á a' a 1 a A a′ A primary stress
ˌa ˊa à ˌa

ˌa (a.) a 2 a a secondary stress
a a a a 0 a8 a a tertiary stress

Title abbreviations


  • ^1 The more precise IPA symbol [ɹ] is sometimes used for English /r/.
  • ^2 Older editions of The Chambers Dictionary used o͞o for ŭ and o͞o for oo.
  • ^3 Nasalized vowel, as in the French phrase un bon vin blanc (IPA: [œ̃ bɔ̃ vɛ̃ blɑ̃]).
  • ^4 Older editions of the Concise Oxford Dictionary used a mix of two systems: the "phonetic scheme" shown in the table above and a system "without respelling". The latter added diacritics to conventional spellings.
  • ^5 These are the stress marks in the print edition. Online, primary and secondary are both written '.
  • ^6 Spelt eye as a syllable of its own; with a consonant, it is spelled y: iodine EYE-ə-dyn; item YT-əm; pipe PYP.
  • ^7 Spelt ə, and sometimes i.
  • ^8 For tertiary stress, not only are the letters italicized, but they are in a different font, also. Secondary/tertiary stress is only marked when judged to be unpredictable, but is not distinguished from primary stress when it is marked.
  • ^9 In IPA, it is an "opentail G" (ɡ / Opentail g.svg, U+0261 ɡ LATIN SMALL LETTER SCRIPT G)[4], not a "looptail G" (g / Looptail g.svg, U+0067 g LATIN SMALL LETTER G)[5].

Pronunciation without respelling

Some dictionaries indicate hyphenation and syllabic stress in the headword. A few have even used diacritics to show pronunciation "without respelling" in the headwords.

The Concise Oxford Dictionary, 1st through 4th edition, used a mix of two systems. Some editions of Webster's Unabridged Dictionary have offered a method for teachers to indicate pronunciation without respelling as a supplement to the respelling scheme used in the dictionary. Pronunciation without respelling is also sometimes used in texts with lots of unusual words, such as Bibles, when it is desirable to show the received pronunciation. These will often be more exhaustive than dictionary respelling keys because all possible digraphs or readings need to have a unique spelling.

Concise Oxford Dictionary's system without respelling
COD variant IPA
ph /f/
kn (initial) /n/
wr (initial) /r/
g, dg /dʒ/ (before e, i, y)
/ɡ/ otherwise
(hard and soft g)
c /s/ (before e, i, y)
/k/ otherwise
(hard and soft c)
ai, ay /eɪ/
air /ɛər/
ae, ea, ee, i.e. /iː/
ė, i.e. (final), ey /ɪ/
ear, eer, ier /ɪər/
aw /ɔː/
oy /ɔɪ/
ou /aʊ/
i͡r, u͡r /ɜr/
eu, ew /juː/

International Phonetic Alphabet

The International Phonetic Alphabet is a standardized method of phonetic transcription developed by a group of English and French language teachers in 1888. In the beginning, only specialized pronunciation dictionaries for linguists used it, for example, the English Pronouncing Dictionary edited by Daniel Jones (EPD, 1917). The IPA was used by English teachers as well, and started to appear in popular dictionaries for learners of English as a foreign language, such as the Oxford Advanced Learner's Dictionary (1948), and Longman Dictionary of Contemporary English (1978).

IPA is very flexible, allowing for a wide variety of transcriptions between broad phonemic transcriptions which describe the significant units of meaning in language, and phonetic transcriptions which may indicate every nuance of sound in detail.

The IPA transcription conventions used in the first twelve editions of the EPD was relatively simple, using a quantitative system indicating vowel length using a colon, and requiring the reader to infer other vowel qualities. Many phoneticians preferred a qualitative system, which used different symbols to indicate vowel timbre and colour. A.C. Gimson introduced a quantitative-qualitative IPA notation system when he took over editorship of the EPD (13th edition, 1967); and by the 1990s, the Gimson system had become a de facto standard for phonetic notation of British Received Pronunciation (RP).

Short and long vowels in various IPA schemes for RP
word quant. qual. Gimson
rid rid rɪd rɪd
reed riːd rid riːd
cod kɔd kɒd kɒd
cord kɔːd kɔd kɔːd

The first native (not learner's) English dictionary using IPA may have been the Collins English Dictionary (1979), and others followed suit. The Oxford English Dictionary, 2nd edition (OED2, 1989) used IPA, transcribed letter-for-letter from entries in the first edition, which had been noted in a scheme by the original editor, James Murray.

While IPA has not been adopted by popular dictionaries in the United States, there is a demand for learner's dictionaries which provide both British and American English pronunciation. Some dictionaries, such as the Cambridge English Pronouncing Dictionary and the Longman Dictionary of Contemporary English provide a separate transcription for each.

British and American English dialects have a similar set of phonemes, but some are pronounced differently; in technical parlance, they consist of different phones. Although developed for RP, the Gimson system being phonemic, it is not far from much of General American pronunciation as well. A number of recent dictionaries, such as the Collins COBUILD Advanced Learner's English Dictionary, add a few non-phonemic symbols /ʳ i u ᵊl ᵊn/ to represent both RP and General American pronunciation in a single IPA transcription.

Adaptations of the Gimson system for American English
/ɒ/ Pronounced [ɑː] in General American.
/e/ In American English falls between [e] and [æ] (sometimes transcribed /ɛ/)
/əu/ This traditional transcription is probably more accurately replaced by /ou/ in American English.
/r/ Regular r is always pronounced
/ʳ/ Superscript r is only pronounced in rhotic dialects, such as General American, or when followed by a vowel (for example adding a suffix to change dear into dearest)
/i/ Medium i can be pronounced [ɪ] or [iː], depending on the dialect
/ɔː/ Many Americans pronounce /ɔː/ the same as /ɒ/ ([ɑː])
/ᵊl/ Syllabic l, sometimes transcribed /l/ or /əl/
/ᵊn/ Syllabic n, sometimes transcribed /n/ or /ən/

Clive Upton updated the Gimson scheme, changing the symbols used for five vowels. He served as pronunciation consultant for the influential Concise Oxford English Dictionary, which adopted this scheme in its ninth edition (1995). Upton's reform is controversial: it reflects changing pronunciation, but critics say it represents a narrower regional accent, and abandons parallelism with American and Australian English. In addition, the phonetician John C Wells said that he could not understand why Upton had altered the presentation of price to prʌɪs.[8]

Upton outlined his reasons for the transcription in a chapter of A Handbook of Varieties of English. He said that the PRICE-vowel represented how the starting point could be anything from centralised front to centralised back.[9] The change in the NURSE vowel was intended as a simplification as well as a reflection that nɜːs was not the only possible realisation in RP.[10] The other alterations were intended to reflect changes that have occurred over time.

Upton's reform
word Gimson Upton
bet bet bɛt
bat bæt bat
nurse nɜːs nəːs
square skweə skwɛː
price praɪs prʌɪs

The in-progress 3rd edition of the Oxford English Dictionary uses Upton's scheme for representing British pronunciations. For American pronunciations it uses an IPA-based scheme devised by Prof. William Kretzschmar of the University of Georgia.


Dictionaries for English-language learners

For many English language learners, particularly learners without easy Internet access, dictionary pronunciation respelling are the only source of pronunciation information for most new words. Which respelling systems are best for such learners has been a matter of debate.

One important difference between the traditional respelling systems and the IPA is that the former contain information about the English writing system and the latter does not. Traditional respellings help learners to generalize about the regularities of English spelling. For example, the traditional respelling of read, past tense, is "red", which exemplifies a common spelling pattern. On the other hand, because it uses different symbols, the IPA transcription rɛd would add to an English language learner's vocabulary without adding information about English spelling. A learner would not know from the IPA respelling that "ea" is less common as a central vowel spelling for ɛ than "e".

The fact that the word for the activity of reading cannot be read without grammatical knowledge indicates the extent of the difficulty posed by English spelling for English-language learners as well as the value for such learners of dictionaries that include pronunciation respellings.

In countries where the local languages are written in non-Latin, phonemic orthographies, various other respelling systems have been used. In India, for example, many English bilingual dictionaries provide pronunciation respellings in the local orthography. This is the case for several Indian languages, including Hindi, Urdu, Malayalam, and Tamil. To reduce the potential distortions of bilingual phonemic transcription, some dictionaries add English letters to the local-script respellings to represent sounds not specified in the local script. For example, in English-Tamil dictionaries, the sounds /b/ and /z/ need to be specified, as in this respelling of busy: "bபிzஸி".[18]

Because these respellings primarily use symbols already known to anyone with minimal literacy in the local language, they are more practical to use in such contexts than the IPA or the Latin respelling systems with diacritics. Another advantage of local-script respellings for English learners is that they retain the "flavour" of local English speech, allowing learners to make connections between their spoken and written English experiences. However, these systems also have limitations. One limitation is that they do not illuminate the English writing system. Like the IPA, they represent phonemes differently from the ways in which the phonemes are normally spelled. So these notations do not guide readers to infer the regularities of English spelling. Also, the practicality of these systems for learning English locally may be offset by difficulties in communication that could arise in the context of other pronunciation norms such as GA or RP.

Children's dictionaries

Most beginner dictionaries are picture dictionaries, or word books. For preliterate native speakers of a language, the pictures in these dictionaries both define the entry words and are the "keys" to their pronunciation. Respellings for English begin to appear in dictionaries for novice readers. Generally, US-based dictionaries contain pronunciation information for all headwords, while UK-based dictionaries provide pronunciation information only for unusual (e.g., ache) or ambiguously spelled (e.g., bow) words.[19]

As the normal age of literacy acquisition varies among languages, so do the age-range designations of children's books. Generally, age ranges for young children's books in English lag behind those of languages with phonemic orthographies by about a year. This corresponds to the slow pace of literacy acquisition among English speakers as compared to speakers of languages with phonemic orthographies, such as Italian.[20] Italian children are expected to learn to read within the first year of elementary school, whereas English speaking children are expected to read by the end of 3rd grade. Pronunciation respellings begin to appear in dictionaries for children in 3rd grade and up.

There seems to be very little research on which respelling systems are most useful for children, apart from two small studies done in the 1980s and 1990s. Both studies were limited to traditional respelling systems without diacritics (setting aside both the IPA and the Webster-based systems used in American dictionaries). Both studies found that in such systems, word respellings may be cumbersome and ambiguous, as in this respelling of psychology: "suy-kol-uh-jee".

The authors of the two studies proposed alternative systems, though there were no follow-up studies. Yule's "cut system" leaves out extra letters, adds specific spellings for sounds with variable spellings, and adds accents to show long vowels, as in this respelling of occasion: o-cà-zhon.[21] Fraser advocated a "non-phonemic" approach using a small set of common spelling patterns in which words would be respelled chunk by chunk, rather than phoneme by phoneme, as in this respelling of persiflage: per-sif-large.[22] According to both authors, the reduced vowel (schwa) need not be shown in a respelling so long as syllabification and syllable stress are shown.

The following overlapping issues concerning pronunciation respelling in children's dictionaries were directly raised by Yule and Fraser: the level of difficulty, the type of notation, the degree of divergence from regular spelling, and pronunciation norms. Yule also raised the question of the types of impact respelling systems could have on children's literacy acquisition. These issues could be usefully addressed in studies that include American respelling systems as well as the IPA.

An issue that has arisen since the Yule and Fraser studies concerns the utility of pronunciation respellings given the availability of audio pronunciations in online dictionaries. Currently the advantage of written respellings is that they may be read phoneme by phoneme, in parallel to the way novice readers are taught to "stretch out" words to hear all the sounds they contain, while the audio pronunciations are given only as whole words spoken in real time. When audio pronunciations are made flexible, it will become possible to study and compare the utility of different combinations of pronunciation features in the online children's dictionaries.

Other uses

Outside of dictionaries, press agencies in English, such as the Voice of America, periodically release lists of respelled given names of internationally relevant people, in order to help news TV and radio announcers and spokespersons to pronounce them as closely as possible to their original languages.

See also


  1. Fraser 1997, p. 182
  2. Landau 2001, p. 121
  3. Merriam-Webster Online n.d.
  4. Landau 2001, p. 118
  5. Landau 2001, pp. 119–21
  6. Such as The Oxford BBC Guide to Pronunciation Mind your language, by Dot Wordsworth, in The Spectator, November 7, 2007.
  7. Lua error in package.lua at line 80: module 'strict' not found.
  9. Upton 2004, p. 225
  10. Upton 2004, p. 224
  11. 11.0 11.1 Lua error in package.lua at line 80: module 'strict' not found.
  12. Lua error in package.lua at line 80: module 'strict' not found.
  13. Lua error in package.lua at line 80: module 'strict' not found.
  14. Lua error in package.lua at line 80: module 'strict' not found.
  15. Lua error in package.lua at line 80: module 'strict' not found.
  16. 16.0 16.1 Lua error in package.lua at line 80: module 'strict' not found.
  17. Lua error in package.lua at line 80: module 'strict' not found.
  18. Students' Deluxe Dictionary n.d.
  19. Oxford Junior Illustrated Dictionary
  20. Seymour 2003
  21. Yule 1991
  22. Fraser 1996


  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Wells, John (2001). "IPA transcription systems for English", at University College London Department of Phonetics and Linguistics site. Retrieved 2006-08-16.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • Lua error in package.lua at line 80: module 'strict' not found.
  • "Introduction to phonetic transcription", at Retrieved 2006-08-16.
  • Oxford English Dictionary. "Pronunciation", from the Preface to the Third Edition. Retrieved 2006-09-10.
  • Oxford English Dictionary. "Key for Second Edition Entries", from the OED website. Retrieved 2014-11-21.
  • Oxford English Dictionary. "Key to New Edition Entries", from the OED website. Retrieved 2014-11-21.
  • Merriam-Webster Online, n.d. "Pronunciation Overview", Retrieved 2011-07-23.

External links