I guess the point i was trying to make, while dismissing the "validity" of it is that encapsulated factual models all still require some references- at some point truth breaks down into trust in the chains of composition.
I dont understand fuzzy logic nearly enough - but it needs some kind of rule set and knowledgebase yes?
Lets say you program a fuzzy logic controller that returns large=true for numbers > 100, and i program mine to return large = true for numbers > 10. Like you say, its subjective.
These rulesets are chosen as truth in the same way the grenwich meantime is 0* lat, and we dont use the paris meantime because we arent french revolutionaries and also the brits had stronger empire.
If i came off as joking its bc i am aware of the lunacy of my take. To call "normal" good is a strong position, but who was norm?? And why do we all have to measure up to him?