Getting softsyn to work

[Will Beldman]

Trying to get text to speech working properly. I downloaded and
installed kernel-2.4.27-speakup using apt-get. As a result,
grep -i speakup /boot/config-2.4.27-speakup
shows this:
CONFIG_SPEAKUP=y
CONFIG_SPEAKUP_ACNTSA=y
CONFIG_SPEAKUP_ACNTPC=y
CONFIG_SPEAKUP_APOLLO=y
CONFIG_SPEAKUP_AUDPTR=y
CONFIG_SPEAKUP_BNS=y
CONFIG_SPEAKUP_DECTLK=y
CONFIG_SPEAKUP_DECEXT=y
CONFIG_SPEAKUP_DECPC=m
CONFIG_SPEAKUP_DTLK=y
CONFIG_SPEAKUP_KEYPC=y
CONFIG_SPEAKUP_LTLK=y
CONFIG_SPEAKUP_SFTSYN=y
CONFIG_SPEAKUP_SPKOUT=y
CONFIG_SPEAKUP_TXPRT=y
CONFIG_SPEAKUP_DEFAULT="none"

and menu.lst in /boot/grub says:
...
title           Debian GNU/Linux, kernel 2.4.27-speakup
root            (hd0,0)
kernel          /boot/vmlinuz-2.4.27-speakup root=/dev/hda1 ro
initrd          /boot/initrd.img-2.4.27-speakup
savedefault
boot
...

And I cannot get sftsyn to work here. /proc/speakup/synth_name says
none and if I try to do
echo sftsyn > /proc/speakup/synth_name
it says speakup_sftsyn module is loaded but if I do
cat /proc/speakup/synth_name
it still says none.

spd-say works fine. speech-dispatcher and speechd_up are installed.
Since sftsyn is compiled right into the kernel as the config file
suggests, how might I go about getting my system to use sftsyn so I
can get all data written to the console read off through the speakers?
According to the config file, it seems to want to use "none". Do I
have to download the kernel source and build from scratch?
-- 
"The unexamined life is not worth living" - Socrates