first let me give you some background on what I am trying to accomplish, I am building a system for an autistic child to use at school , she has no speech , so I have a small computer with common phrases in MP3 format so she can make requests of her teacher , what I am trying to do is edit the initrc file to map a key to play the file , without alot of success I might add , what I did was map "h: to bring mpg123 up and play the sound clip , ("h":mpg123 --gain 100 happy.mp3/C-M") seems to work until I start adding lines to map the keys further ("s":mpg123 --gain 100 sad.mp3/C-M") any help would be appreciated , both by me and this child , I would really like to give her a voice of her own !

Look up text to speech synthesizers. You can use speechd which creates a device node so you just echo text and the computer will speek. speechd uses Festival or rsynth. Festival will give better quality.

To replicate her voice takes a lot of research of her throat and vocal cords.

the problem is that the child we are doing this for has the intelect of about a 3yr old , she is not Autistic in the strictest scence but suffers from RTS a rare syndrome simalar to DOwns (sp) syndrome , she will never advance more than a 4 yr old and will probably never speak.
I think I have the mapping feature figured out , the problem was staring me in the face , I was mapping h to a command and the command contained an h causing it to fall into a loop, I am using DSL , only problem now is that the volume is too low at startup , it has to be adjusted to 100% in Dmix , looking for the init file now that controls it so I can set the default start to 100%