IPA is not known by everyone, especially when the site is aimed for learners. It is expected that people here are likely to be ignorant about it (e.g. myself). Therefore, I suggest contributors here use less IPA in their questions/answers so that everybody can understand the posts. For example, I found this answer amazing and a pleasure to read. The answerer added IPA, but as a side note. I enjoyed reading the answer as I could understand it well. So I request people here to use the English alphabet more to clarify the pronunciations. However, IPA can be added as a side note.

  • 2
    What do you mean by “alphabets” here? I do not think that that word means what you think it means.
    – tchrist
    Commented Feb 8, 2013 at 2:21
  • @tchrist, "a","b","c",...... etc. If not alphabet what should have I said then?
    – Mistu4u
    Commented Feb 8, 2013 at 4:50
  • @tchrist elL, alphabets shouldn't be plural, but there's nothing wrong with the word choice.
    – mcalex
    Commented Feb 8, 2013 at 9:51
  • 2
    @mcalex You should have said letters; you’ve listed letters, not “alphabets”. An alphabet is an entire (usually ordered) set of letters used for writing a particular script, such as the Greek alphabet, the Latin (=Roman) alphabet, the Cyrillic alphabet, etc. Computer programmers may sometimes call letters “alphabetics”, although just plain letter is almost always better. You’ll also hear “alphanumerics” in computer contexts, which means letters or numbers, and almost always numbers restricted to decimal digits. English uses the Latin alphabet, although it has added its own letters.
    – tchrist
    Commented Feb 8, 2013 at 13:36
  • @mcalex I disagree: see his comment, and mine.
    – tchrist
    Commented Feb 8, 2013 at 13:37
  • @tchrist I'm guessing your first comment to me is to Mist? :) To the discussion: all the letters (a, b, c ... etc) == the alphabet. He's saying he wants the IPA alphabet (or all the IPA letters if you prefer) used less predominantly.
    – mcalex
    Commented Feb 8, 2013 at 14:40

5 Answers 5


I agree with @tchrist: IPA is a response to phonetic ambiguity. A scientific response.

But I agree with you, too: a language learner is less likely aware about IPA (than a linguist).

Please take a look at this discussion. Your question seems to be a subset of that one.
The answers from there suggest a dual approach:

  1. Do not throw away terminology just for sake of simplicity.
  2. Always expand the meaning so that an average ELL grasped it quickly and started using in their future study.

Personally, I would place IPA first, but that's negotiable.


You cannot explain how something is pronounced using the normal letters used in English, because the mapping between spelling and pronunciation in English is extremely complex and even idiosyncratic.

That’s why the standard International Phonetic Alphabet was invented: so there can be a standard way of indicating pronunciation. You cannot do that using simple A–Z.

  • Further, there are some permissible phoneme sequences that cannot be unambiguously notated using an ad-hoc spelling system. Either we use the IPA, or some formal transcription system (Wikipedia has one, based on one of the dictionaries'). Commented Feb 7, 2013 at 11:24
  • 1
    You can explain it with A-Z, it's just you cannot do so succinctly. For instance, "phonetics" can be described as "'ph'. which sounds like 'f' in fox, 'on' as in 'only', 'et' as in 'net', 'ics' as in 'bricks'". This is somewhat less succinct than /fəˈnɛtɪks/ though.
    – Matt
    Commented Feb 10, 2013 at 4:01
  • 4
    @Matt That doesn’t work for most words, because not everyone says them the same. Even here you got it wrong, since there is no schwa in the stressed syllable of only, as you allege to occur in the first syllable of only, It really does not work.
    – tchrist
    Commented Feb 10, 2013 at 13:28
  • Each word would have many, many IPA transliterations. How many English accents are there!?
    – jsj
    Commented Feb 11, 2013 at 6:30

I don't entirely disagree with other answers. However, I can't read IPA without a dictionary/google in front of me (I certainly wouldn't try writing it - dictionary or otherwise), and I am a native speaker who has a fairly good understanding of the language. (Mental note: pleeease try not to forget to check spelling. :-)

In addition, regardless of the how the IPA says to pronounce a word, there are so many different accents in so many different English speaking countries that natives using one will disagree with different natives using another. English is not a scientific language. OTOH, grass rhymes with glass whether it's pronounced with a long or short 'a' sound, so using one to assist with pronouncing the other works whichever way you come at it.

Don't get me wrong, I think answers with added IPA information - as per the OP example - are excellent and deserve upvoting. I've also come to the realisation that I'd better start learning IPA if I'm going to remain hanging around. :-) My concern is that answers won't be deemed as showing much effort if the IPA info isn't there - much as somebody referencing Wikipedia would (correctly) be called out for not including a link to the page. I don't think people unaware of IPA should be downvoted for not including it in an answer.


Even among people who know IPA there is great variation in how they use it, and worse, many of them are ignorant of the variation and believe there is only one correct way to us it.

Of course this problem also exists for ad-hoc transliterations "spelling in phonetic English".

For instance how to indicate the first vowel in "ago", the vowel in "slur" or the vowel in "roof" and is it the same as the vowel in "book" or the sound in "boot"? In the end you either leave some people confused or have to come up with an actual system that everybody has to learn to use correctly.

Pragmatically I would say:

  • Sometimes we can render the sound of a word clear for all who know English without needing a system.
  • Sometimes if we are not careful in rendering the sound of a word "in normal English letters" it will represent different sounds to different people. To avoid this we would need to choose one system and stick to it.
  • It should never be "wrong" to include IPA alongside a "normal English letters" rendering of the sound of an English term. IPA has fewer ambgiuities so those who know it will be able to edit the "plain English letters" in keeping with it when the plain letters turn out to have an interpretation other than what the writer intended.
  • So do you indicate IPA is not free of errors even though it is "a scientific response" as per an answer here. I think your point 3 indicates that.
    – Mistu4u
    Commented Feb 10, 2013 at 9:51
  • "Errors" is not the best word to describe any shortcomings or weaknesses in the IPA, as for how people use it sometimes errors are made and sometimes there are more subtle problems mostly due to misunderstandings. Commented Feb 10, 2013 at 10:06

Unfortunately, there is no standard representation of English phonemes.

IPA is widely used among linguists; but it is not really satisfactory, since it represents pronunciations rather than phonemes. It implies that some phonetic realizations are preferred to others, and it draws distinctions which are not necessarily perceived by hearers. It is, moreover, ethnocentric; it employs the Roman alphabet in a manner which is (relatively) easy for speakers of most European languages to grasp, but is bewildering to those whose principal knowledge of that alphabet derives from the study of English.

Over on ELU, John Lawler has endorsed this scheme; but that is still dependent on non-Roman-alphabet characters.

I think the best course is that which OP and mcalex recommend: to describe pronunciation using ordinary English spellings, and annotate with broad-transcription IPA. That has its own pitfalls—how, for instance, does one regularly distinguish /æ/ in ‘ordinary English spelling’?—but I really don't see a better way.

  • 1
    This is not true. IPA can be and is used for both phonetic and phonemic transcriptions. The custom is to wrap the former in square brackets [thus] and the latter in a pair of forward slashes /thus/. Commented Feb 10, 2013 at 9:32
  • @hippietrail Yes, IPA can be used for phonemic representation, but it is not designed for it. How, for instance, do you represent 'oh' - as /əʊ/ or /oʊ/ or /o/? What glyph do you use to represent 'R'? ... It is the same problem which you very cogently raised against Wiktionary's favouring one spelling over another. Commented Feb 10, 2013 at 12:45
  • Please provide some corroborating evidence for your claim that IPA is not designed for phonemic representation. The vast majority of English dictionaries published outside the US, including Oxford, have used the IPA for phonemic representation this way for decades. Commented Feb 10, 2013 at 12:51
  • It's a very common misunderstanding that there ought to be one single correct way to use IPA to represent phonemes of a single language. If you look in any dictionary which uses IPA you will find a pronunciation guide telling you which symbol is used for each sound, just as you will find in dictionaries using a pronunciation system not based on IPA. This is indeed a weakness in using the IPA alone on this site. Personally I prefer /əʊ/ for "oh" in British and Australian pronunciation and /oʊ/ for it in American. I avoid /o/ for English. For "r" I prefer /r/ in most cases. Commented Feb 10, 2013 at 12:58
  • @hippietrail And syllable-terminant 'R'?! I think in fact we're in agreement. Two critical questions you have raised (how to avoid favouring one pronunciation, here, and the status of the phoneme on linguistics.SE) are united in the practical question What can we usefully tell learners who come from and are seeking to enter very different speech communities? Phonemic representation is essential; but Whose phonemes? and How shall we represent them? Commented Feb 10, 2013 at 13:28
  • Yes @StoneyB this is a big question! Like I said we will surely have to adopt a system and have a pronunciation guide, even if we have a "plain English system". But we might only need to use it for ambiguous cases and give leeway when there's no ambiguity since most people here will be beginners. For syllable-final "r" in American IPA there a couple of special "hooked" vowels, including schwa. I'm not sure what's done in other rhotic English varieties like Irish English. British and Australian English have a "joining r" that's only pronounced when followed by a vowel. For this I use (r). Commented Feb 11, 2013 at 3:17

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .