Scheidt implies an interest in duress ciphers that is relevant to Kryptos

Ok, so at first I didn’t know what duress ciphers were.  I thought they were literally a type of cipher that had two different messages with two different keywords or something.  That is NOT a duress cipher.  A duress cipher is one that has been sent under conditions of duress that contains pre-agreed upon signifiers of duress.

Merriam-Webster on duress:

1: forcible restraint or restriction
2: compulsion by threat
The idea is old and use was common in WWII, although from the war stories not always successful.  The basic idea is that if you get caught there are keywords or references that can be used as a type of code to signify duress or also more steganographic means such as misspellings.  It should be pretty obvious now what the misspellings in Kryptos likely signify.  Now keep in mind that it’s not the presence or absence of misspellings but a pre-agreed upon signal.  In Kryptos, it’s probably not that there are misspellings but the characteristics of those misspellings (i.e. location).  The reason duress ciphers are useless is that the captors would often have access to previous transmissions and simple comparisons would reveal patterns.  Also, they would usually beat everything out of captives which would include any type of duress signalling.  It’s a valid idea but probably only realistically likely between close friends on either side of the message with access to a slew of inside jokes or past contextual references that could survive efforts to turn the captive or use the rubber hoses.
Duress ciphers are not a magic combination of dual keywords that consistently enable the users to send different messages depending on duress or normal conditions.  You can artificially jury-rig examples to force this to work but neither I nor anyone I’ve yet seen have come up with methods simple and reliable enough to use that are also secure.  At this point, duress will simply be an adjective describing the environment the message was sent from.