Computerized Phonemic Speech

You know computerized speech? Where you type in something and it "speaks" it for you. Sounds all stupid. You hear it all the time in janky electronic music.

Why haven't we improved that technology in over a decade?

Here's my proposal:

Instead of typing in "Hello, I'm a cheap computer" orthographically, use the IPA system: "hɛloʊ aɪm ʌ tʃip kəmpjutɚ".

THEN, oo, this will be good: you can DRAW the suprasegmentals right over the letters:
volume, pitch, even rate. Drawn on. Kinda like in a MIDI sequencer or something.

The letters could shrink or increase in size as you adjust the volume up or down. They could lengthen or contract as you adjust the speed faster or slower. Pitch could be the customary line riding above the letters.

Sure, it will be more complicated from the user's end. But also way more fun. And you'll get a much more human-sounding result.

Let's pretend that this technology were widely available, and some nerdy-yet-sizable percentage of the general public got more-or-less used to typing this way.

  • Internet content typed this way would be immediately available as audio. (Imagine Wikipedia pages that you could listen to on your mp3 player!)

A program could be written that would automate the process; it would turn orthographic text into phonemic text. Ideally, the results would be edited by a human to check for things like homographs ("read" past tense, "read" present tense). But actually, maybe not so much if the program had a built-in grammar checker to choose the proper homograph based on context. Potentially, the human editor could also add the prosity, dramatic pauses and length. But all that are the nerdy details, which wouldn't really matter in the face of the big picture: Any text that has been digitized can be made instantly available in audio format. Actual listenable audio format. Not janky robot format.

Even Bigger Picture:
If this technology catches on, future layman will know the IPA alphabet and use it widely throughout the internet. Even keyboards would be updated.

  • People studying foreign languages could do so with much more attention to pronunciation.
  • The blind would have access to all digitized text
  • People with apraxia of speech or other communication deficits (i.e. Stephen Hawking) could communicate in a much more normal-sounding way.

Boo YA!

No comments: