The problem that I see with this is that people are basically lazy-brained. Even if a language requires that you choose a final particle that indicates evidentiality, people will just not use it. For example, if the written form of a language requires that a sentence end with ".", "?" or "!", and each one is an evidential particle, then tomorrow someone on the internet will say "By the way everyone, I'm tired of doing all of this evidentiality stuff when I don't need to, so I'm just going to write '_' at the end of all of my sentences, and it doesn't mean anything but that the sentence is over." Within a week the convention will be adopted all over the world, and mandatory evidentiality will be a thing of the past. It might or might not be a good idea, but I just can't see a grammatical requirement overcoming human laziness.
Many years ago an acquaintance of mine in college said "A system without an application is a useless ornament." I believe that he was quoting someone, although I have no idea who (BTW, if anyone here knows where this quote might have come from I'd appreciate the reference). In the case of a conlang, part of the beauty comes from the fact of its widespread use. While I agree that mandatory evidentials are a tempting idea (I'd certainly like a language that has them), I don't believe that they'd hold up well in actual use.
For example, if the written form of a language requires that a sentence end with ".", "?" or "!", and each one is an evidential particle, then tomorrow someone on the internet will say "By the way everyone, I'm tired of doing all of this evidentiality stuff when I don't need to, so I'm just going to write '_' at the end of all of my sentences, and it doesn't mean anything but that the sentence is over."
Actually ".", "?" and "!" are illoctutionary operators. Sentence have a different ...
I'm working on a conlang (constructed language) and would like some input from the Less Wrong community. One of the goals is to investigate the old Sapir-Whorf hypothesis regarding language affecting cognition. Does anyone here have any ideas regarding linguistic mechanisms that would encourage more rational thinking, apart from those that are present in the oft-discussed conlangs e-prime, loglan, and its offshoot lojban? Or perhaps mechanisms that are used in one of those conlangs, but might be buried too deeply for a person such as myself, who only has superficial knowledge about them, to have recognized? Any input is welcomed, from other conlangs to crazy ideas.