Posts Tagged ‘phonetizing’

How to transcribe English into phonetic alphabet using

2012/07/05 2 comments


Phonetizer transcribes into IPA. The vocabulary seems somewhat limited (45000 claimed) – English spelling variants do not help, although Phonetizer offers BE as an input option. I have not found a length limit for the transcription with an article from the current Economist of over 1000 words – should be plenty for most reading/recording assignments in the LRC. Easy as (web2)py. Smile  The web version is advertisement-based. The downloadable version is not free, so we cannot install it in the LRC, unfortunately.

Google-Translate for phonetization?

        1. Google-Translate also offers some phonetic transliterations. You may have noticed this when attempting (remember, though, that it is for a reason that they link to “professional translation” services, and also invite anybody to amend the machine translation offered) to translate from English into other languages, image
        2. However, if you type or paste non-Romanized text into the source textbox, you also get the option button “read phonetically” (meaning transliterate to phonetic symbols or phonetize). image
        3. Limited use in the LRC: Few languages are supported.
          1. Only languages written in non-roman letters are offered. E.g. French or German are not deemed difficult enough (I know a few that would beg to differ Smile).image
          2. Arabic, Farsi and Hebrew are also not supported (root cause: right-to-left? Strangely these right-to-left-languages work in the TBA:Google transliterate IME which attempts to do roughly the opposite of phonetization): image
          3. Leaves: Chinese, Greek, Hindi, Japanese, Russian. However, note finally that not a standard phonetic alphabet is being used either for these transcriptions.

Phonetic transcription websites

Computerized Language resource centers are supposed to work wonders improving SLA students’ pronunciation: Can’t computers analyze and visualize sound for us?

However, it turned out there seems to be a considerable “impedance mismatch” not only between computers analyzing and understanding the signal, but also between a computer voice graph and the capability of a language learner to process and improve pronunciation on the basis of it.

Voice graphs may have some use for tonal languages. But can you even tell from a voice graph of a letter  which sound is being produced?

Enter the traditional phonetic transcription that pre-computerized language learners remember from their paper dictionaries (provided you can teach your language learners phonetic symbol sets like the IPA). Not only are good online dictionaries perfectible capable of displaying phonetic symbol sets on the web (it’s all in Unicode nowadays). 

There are now experimental  programs that can automate the transcription of text into phonetic symbol sets for e.g. English, Portuguese or Spanish. The more advanced ones also come with text-to-speech.

You can provide your students with audio (or, text-to-speech capability provided) or text models and have them study the phonetic transcription, listen to the audio, and record their model imitation in the LRC. Maybe you will find that practice with recording and a phonetic transcription of the recorded text is more useful for your students’ pronunciation practice than a fancy voice graph.