Speech recognition

Giles Bowkett argues that every programmer should develop a personal style,
and Andrew Norris responds that this is not a good goal. He seems to think that this is akin to writing Fortran in any language.

I don't quite agree. Personal style should be orthogonal to the style imposed by the idioms of a language. It reminds me of speech recognition, which can have several outputs: First, the words spoken. Second, who's talking. Third, the mood of the speaker. I don't have experienced 'speaker recognition' (that isn't tied to idiom usage) in source code yet, however, and can't quite think how that would look like.

I do remember, however, reading a german book about electronics, whose tone was just slightly unusual. It turned out to be a translation from finnish.