[DECtalk] why can dectalk sing

[Carlos Fernandez]

The process of synthesizing speech involves the computer finding the 
correct phoneme, and playing that phoneme at a pitch and time extent to 
sound natural in the active language. This is actually more involved 
than singing, where the values are manually defined. Thus, any speech 
synthesis system using fully computer-generated samples is capable of 
singing. As dectalk was developed in the 1980s, where manual control was 
more common, and as the result was interesting, the singing ability was 
made easier to operate, and thus it has remained.

Carlos