possible thought for enabling speech recognition

[Eric S. Johansson]

This may sound a bit odd but hear me out

NaturallySpeaking automatically enables Select-and-Say functionality you've you 
use the right edit control.

The simplest way make something edible with speech recognition from Emacs is to 
copy it into a Windows edit control. The edit control window would have its own 
grammar and action routines to operate on it.

For tasks which do not require any information except what can be seen in the 
buffer, this is a possible solution. It's probably also a good solution for any 
task requiring more information than what is visible. Problem being as I 
illustrated earlier communicating data to and from the Emacs environment.

So the question is, can we get the proper edit controls invoked from Emacs?