I read the article and it presented quite complicated thoughts in a language slightly beyond my current english capabilities.
If I understand it correctly its a combination of philosophical and mathematical approach to perception, decisionmaking and something what is referred as a knowledge. It reminds me mehods i applied while I was working on game AI.
The example with zebra looks completely flawed to me. Are we looking at a real animal, or a photograph, or a digital simulation which might change its shape?
One of the approaches i implemented to AI was a question “Do I have enough information to make a decision?” and when answer was “No” the AI was directed to obtain more information about the subject. Logic presented in the example ends for me as “lazy written script” where author decided to end the script without getting additional data - which of course is the more complicated way of making a better decision. If it should be a case against Epistemologic Contextualism, its a bad example. Not sure if it was a “strawman” or somebody really tried to argue against it this way.
Examples which were used for EC look usually logical up to same point, where there is made an “negative” assumption, that something is not as it looks. Like the zebra which is not a zebra. There is an unbased doubt, which is later exagerrated and original claim about zebra is abandoned on assumption.
For the purpose of AI i had to order which information are more or less trustworthy.
a) Information gained first hand, verified more than once, unfalsified
b) information gained from other source, verified more than once, unfalsified
c) Information gained first hand, verified at least once, unfalsified
d) Informaiton gained from other source, verified at least once, unfalsified.
e) Information gained first hand, unverified, unfalsified
f) Information gained from other source, unverified, unfalsified
g) Information gained first hand, but falsified at least once
h) Information gained from other source, but falsified at least once
i) Information gained first hand, but falsified multiple times.
j) Information gained from other source, but falsified multiple times.
Lets say we start at E. We looked upon an animal and “its a zebra”.
Problem with the zebra example is that original decision, already made, was retracted without actual falsification, just when a claim was made. My usual position is that there has to be a reason to doubt, not just the emotional feeling, and next reaction should be verification, not retracting of the decision.
After the process we either end up at G if we found out its a Donkey, or at C when its confirmed its an actual zebra. If the examination isnt performed, we stay at E and original decision should not be retracted. So, in 2 out of 3 cases no change to decision. It explains why are people natively conservative.
Most problems are at information from other sources without verification, or when facts are not taking into consideration, or when falsification is not possible.
In regard to common perception and religion:
Religious people claim they have knowledge at A or sometime at B. in fact they have grade F. (looks high, but it isnt. Anything on grade E or lower isnt worth spreading)
Skeptics usually usually have and discuss information at grade B and D, if they are scientists or debunkers its A and C - you just need to take part in verification and eventual falsification.