Would it not be nice to have a site from which people can listen to different phrases in different languages with each phrase having the following characteristics, in any combination:

  • Sample Sentence (describes the meaning of the sentence)
  • Name of Person Speaking (used to differentiate among speakers)
  • Language and Dialect (defines the written form of the words in the sentence)
  • Accent (local, or foreign and in the latter case original languages(s) and dialects(s) of the speaker, thus leading to mixed accents as well. Most importantly an accent defines not only the sound but also the rhythm of the words and letters therein pronounced within the sentence, again, something that can take some time to get used to.)
  • Intonation (happy, sad, fed up, reassuring, complaining, enthusiastic, surprised, etc...)
  • Sexuality of the speaker (such as, for instance, any combination of not male, male-, male, male+ with not female, female-, female, female+) once again defined in a somewhat flexible manner according to the speaker's self perception.
  • Speaker pitch (soprano, alto, etc...)
  • Other physical voice attributes (e.g. having caught a cold, having a sore throat, etc...)
  • Sentence pronunciation in IPA or other phone/phoneme transliteration system
  • Optionally, a list of sentences which are not the sentence being uttered but that the listener could confuse in place of what is being said.

On this site people could also be allowed to upload their mp3 or wav sound files so as to contribute sample sentences, or conversations, which the user of the site can then listen to at a later time.

Does anyone know of such a site? If not then it could be a good idea to construct one like this? Such a site could be a great aid in allowing learners to practice with different languages, dialects, and accents etc... as described above, one more step towards preserving the linguistic diversity of our people, places, and world while learning about them in a fun, accessible, and easy manner.

  • 4
    Hi John, while this would be fascinating, I think you might have to rephrase this and make it more like a question (have a look at the FAQ). However, if you are interested, then the BBC Voices page is a good place to start: bbc.co.uk/voices Jan 7, 2013 at 14:24
  • This has some of the features you're looking for as well: accent.gmu.edu
    – lapropriu
    Jan 8, 2013 at 1:10
  • I found this paper about building a cross-linguistic English/Mandarin accent corpus. Not entirely surprising that it's from a cell phone company.
    – acattle
    Jan 8, 2013 at 3:08

2 Answers 2


There is a website for accents: http://www.forvo.com/

They don't have sentences. (I think the website would become unmanageable if they did.) I don't know if they have idiomatic phrases, but they have quite a few single words and compound word phrases.

Anyone can contribute by submitting their own samples, and by rating samples provided by other users.

This one, for example, has the word 'bottle' as pronounced by Brits, Americans and Canadians.


  • This is more of a pronunciation guide for learners who want to learn the 'standard' form of a particular language I take it? What is the rating based on? Would my pronunciations be rated badly for not being RP or GE pronunciations? It's an interesting resource, though. Feb 8, 2013 at 16:51
  • @DangerFourpence: I believe the idea is to rate pronunciations based on how representative you think they are of the region they are spoken in. So, you could be speaking Jamaican or Nigerian dialects. As long as you mark the appropriate origin, it serves its purpose.
    – prash
    Feb 9, 2013 at 3:01

There is something fairly close to what you're talking about, which is the Speech Accent Archive someone above mentioned: http://accent.gmu.edu/

This is a collection of speakers from different regions and with different native languages (there are a lot of native English speakers from various regions but also non-native English speakers with various native languages) reading the same short passage. The passage is a little bit weird but I believe it was designed to include as much of the English phonological inventory as possible.

Entries in the Speech Accent Archive include the speaker's birthplace, current place of residence, age at time of recording, native languages, other languages spoken, age at which they began learning English, where they lived that English was spoken, and how long they've lived there.

Some of the passages have been transcribed in IPA, but unfortunately not all. I'm not sure how new transcriptions are added.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.