I don't think either test involved the speaker conveying information about the SCP to the subject. The information was pinned to the wall in pictogram form.
In test 1, the subject was commanded to translate the pictograms into text.
In test 2 the subject was commanded to describe the pictograms verbally.
Re:●●|●●●●●|●●|●
by blamestross on Tuesday May 11, 2021 @05:54PM
(#27124046)
Huh, the server has not been stolen yet, maybe it doesn't work the way we think.
Information given in the SCP tell us that digitally stored files are affected to. That seems to mean digitally stored text.
Text in a digital form is just a way to indirectly reference pictures to be shown on the screen or printed. Sometimes it's a straight lookup table correspondence (ASCII), sometimes more complex (Unicode); there may be noise (OOXML). I'd guess that, in context of this SCP, a PNG with letters on it would count as text too.
This suggests that the issue isn't with a representation format, but the interpretation of it - or at least in a way the representation is processed by human minds.
I put forward a hypothesis that the relevant recognition criteria is words being recognized from sensory inputs by the human brain, and backtracking to most direct source of these words. This would explain the results of both documented tests. I propose following further tests to be made:
- Repeating Test 1, but blindfolding the test subject before telling them to write words down, so that they can't see them. Expected outcome: like in Test 2.
- Repeating Test 2, but instructing the test subject to not repeat the words heard out loud. Expected outcome: like in Test 2.
- Repeating Test 2 with the test subject incapable of hearing. Expected outcome: nothing happens.
- Repeating Test 2, using a machine recording sounds and replaying them after a delay, in place of the test subject. Expected outcome: nothing happens.
- Repeating Test 1, using a computer connected to a microphone and a printer, in place of test subject; the machine would run a dictation algorithm to print the spoken words. Expected outcome: nothing happens.