Oh, really? Name one. (;
Wrong. Normal floating point functions (exp/log/etc) need to fallback to using multiple precision calculations to ensure that the result is perfectly accurate. Much works is done to avoid this having to happen often, but sometimes it is necessary.
This is just a ploy by SERN to divert our attention away from their clandestine plans to take over the world.
It isn't by electronegativity, though. See http://en.wikipedia.org/wiki/T...
Extremely relevant story - http://www.marshallbrain.com/manna1.htm