[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Voice Stress Analysis of Debates? [NOISE]



As far as the technical issue, can you use digital signal processing
to insert/delete the features that the VSE is looking for?
Sure, if you know what the VSE product is looking for.
(And you can crank up the bass while you're at it.)
Doesn't matter if you call it microtremors, it's a characteristic
well-defined enough that an $89.95 retail hardware device can detect,
so it's a characteristic you've got a not-terribly-complex model for.
If all the popular VSEs use the same algorithm, or if you know what
brand the TV Networks were convinced to use, you're fine.
(If you don't know, that makes it tougher, but you can experiment....)
Or you can distribute a cheap software-based VSE program, with hooks that
let you feed it "Truth" or "Lying" overrides :-)

At 11:28 AM 10/7/96 -0800, jim bell <[email protected]> wrote:
>I think it's particularly revealing that the TV networks don't try to use 
>it.  They're supposed to be looking for an edge, something to make the news 
>seem more interesting.  The political establishment would see the 
>publicizing of  this as going beyond an unwritten limitation on the media.

I've never seen the TV networks use handwriting analysis to spice up
their news reporting either, and the only "news" shows I've only seen 
mention people's astrological signs were the Entertainment-Tonight types.
(I'd rate VSE as slightly more scientific than graphology, and both of them
somewhat above palm-reading and astrology, and a bit below polygraphs.*)
They get more market asking what kind of underwear the candidates wear.

Besides, what would they do with it - subtitles "He's Lying" or
"He Seems To _Believe_ This Nonsense?" or "He Doesn't Really Care What 
The Teleprompter's Telling Him To Say"?  You'd need to leave at least
one of them on during the whole Bipartisan Establishment Debate**,
and you don't need any piece of hokey machinery to 
If they actually _wanted_ to make the shows more interesting, the networks
could have reporters ask the candidates real questions.  But they
wouldn't get invited back next election....  And if they ask questions like
"How could you possibly expect us to believe _that_, Mr. President?"***,
they'd get thrown out immediately.

[*VSE and Graphology have some chance of a causal basis, though
polygraphs have the extra advantage that the people administering them
get to intimidate the victim while doing the test, which is more likely
to get results than running VSE on a speech somebody's practiced and
is reading off a teleprompter, even if interpretation of both weren't
pretty much subjective anyway.]

[**And they'd have to interlace the subtitles with the fnords.]

[***Did you ever notice how much George Bush _sneered_ when he was lying?
Clinton's polite enough to smile when he's talking.]


#			Thanks;  Bill
# Bill Stewart, +1-415-442-2215 [email protected]
# You can get PGP outside the US at ftp.ox.ac.uk
  Imagine if three million people voted for somebody they _knew_,
  and the politicians had to count them all.