From: karn@qualcomm.com (Phil Karn)
Date: Wed, 23 Jun 93 00:23:02 PDT
To: cypherpunks@toad.com
Subject: Re: YAA (yet another article)
Message-ID: <9306230722.AA00336@qualcomm.com>
MIME-Version: 1.0
Content-Type: text/plain


At 01:46 PM 6/16/93 -0400, Perry E. Metzger wrote:

>Torture, believe it or not, is a very effective way of police to get
>information. Our society bans it. Every mechanism that is useful is
>not acceptable.

Not to dilute your argument or anything (you know I agree with it), but
the reading I've done on the history of the Fifth Amendment says that one
of the reasons torture was eventually banned in Western countries (through
mechanisms like the American Fifth Amendment) was the growing realization
that it actually was NOT particularly effective. People would falsely
confess to all sorts of crimes just to get the torture to stop. Consider
how many confessed witches were burned in New England.

One of the problematic things about encryption (as it's usually practiced
now) is that it's relatively easy to tell if an encryption key is the right
one or not. This makes it tempting to resort to torture (or a "contempt
of court citation", in modern terms) to extract it from an unwilling
defendant. That's why both steganography and "duress key" schemes will
remain important for some time, even if the 5th amendment were to be held
as applicable to compelling crypto keys. You could cry "torture", while the
police would claim that they discovered the key by other means (or that you
disclosed it "voluntarily") and it would be your word against theirs.

Phil