I am looking for a chart (or good reference) that shows every possible orthographic representation for each phoneme (in General North American English; I don't care about the low-back merger).

Something like

/a/.......au (caught), o (pot), a (garden)

But for each phoneme

  • 1
    There might be some charts out there that attempt to do this, but none will be without issues. If you sit down and attempt to make one yourself, you will quickly see that letter-to-phoneme conversion doesn't always work like that. What letter(s) correspond with the /ay/ sound in the word bide? Is it the <i>? The <i> and the <e>? In the word cache, what makes the /ʃ/ sound? Is it the <che>? What about in cached? Does the <che> correspond with /ʃ/ while <d> corresponds with /t/? But what about cash and cashed? What are the phonemes in rangle? How do they line up with the letters? May 29, 2015 at 12:11

1 Answer 1


I agree with musicallinguist, but you can get this information for a given corpus if you really want to. One option is a tool called Phonetisaurus, which is one of a family of statistical graphone tools. Its intended function is grapheme-to-phoneme conversion, but one of the data files it builds along the way is a set of grapheme-to-phoneme mappings attested in the input corpus, weighted according to the computational model. It attempts to deal with the issues of context/cooccurrence that musicallinguist noted. This would be relatively tricky unless you're pretty familiar with the programming, and probably more effort than you'll want.

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.