In article , Arthur Hagen wrote: :Allow a little pedantry here... You can't trace anyone from fingerprints or :DNA on a piece of evidence. You can use the fingerprint/DNA to obtain a :positive identification with someone on a list of suspects, but it's not :backwards traceable as such
True, but if a company suspected an inside leak and was displeased enough to persue it by hiring investigators, then it would be relatively simple to take samples from (e.g.,) the front-door, mix them together, check to see whether the test sample was fairly close to something in the mix -- and once so determined, to use the standard "detect the counterfeit coin in a mix" algorithms on mixes of samples from (e.g.,) office doorknobs or keyboards to narrow down the search area drastically.
:> As a simple example: even if I was careful to run my messages through :> a spell checker, it would be clear to any practiced analyst that I am :> Canadian, if if only by examining my pattern of choice of 's' versus :> 'z'.
:Most people have idiolects that can betray them, but knowing that can help :you avoid it.
Of course, I would know to use (e.g.,) color instead of colour . 'o' vs 'ou' is easy -- but 's' vs 'z' is much less rational and the alternative spellings really *look* wrong.
: Since I know I tend to write fairly long sentences and use :certain words and phrases more than what's common, I can try to avoid doing :that. It won't be foolproof, but enough to not immediately point the finger :at myself.
I suspect I would have to babblefish through half a dozen languages to scrub my idioms adequately ;-)