As far as I know, most constructed languages have no syntactically or semantically ambiguous grammatical constructions. However, would there be any reasons that a constructed language might include certain potentially ambiguous constructions, and if not, why do many (or maybe most) natural languages have so much ambiguity, and why would that not be resolved in natural languages?
It's a hard to intentionally change a natural language. This is why there is ambiguity--because they have evolved in that way. So, having ambiguity in a conlang can make it seem more natural– CHEESEFeb 8, 2018 at 14:23
4"most constructed languages have no syntactically or semantically ambiguous grammatical constructions" This is most definitely not true. For example, many conlangs use a single word as a relativizer (as in "the apple that I ate"), which is certainly cause for ambiguity (there may be multiple potential locations for the trace). Generally, it would be very difficult to create a naturalistic grammar where every sentence only has a single parse (barring Lojbanic clause-terminators that basically amount to spoken punctuation).– DoorknobFeb 8, 2018 at 15:07
If naturalness (being like a natural language) is a design goal, then a conlang can embrace ambiguity with no shame.
Efficiency. Most ambiguous sentences are understood well enough in the context they are uttered in, clarified either by prior knowledge, non-linguistic communication (body language, pointing, etc), or through follow up questions from the listener. When most ambiguity is not actually a problem, it doesn't need to be explicitly countered. To require a language to be completely and always unambiguous would require much more specific and cumbersome sentences, violating the cooperative principle.
Wordplay. Puns, double entendres, and garden path sentences are unlikely to be possible or productive without substantial ambiguity.
1Regarding point 1, ambiguity would even be required, since as OP said, natural languages tend to have such ambiguity.– MasclinsFeb 9, 2018 at 10:11
Garden path sentences aren't really necessary for a language, imo -- they're more of an interesting way to look at how we process language. A better example for point 3 would be things like Gricean reasoning, politeness, and other pragmatic phenomena, which often rely on ambiguity or at the very least vagueness. Feb 11, 2018 at 23:28
@Sparksbet This question doesn't ask about necessity, just why you'd want ambiguity. Not all conlangers want wordplay, but if they do, then I think that implies ambiguity. I'm not sure what Gricean reasoning is, though Grice's maxims are part of the cooperative principle, maybe politeness too?– curiousdannii ♦Feb 12, 2018 at 1:56
1@curiousdannii by Gricean reasoning I meant those maxims and how you derive conversational implicature from them. You mention the cooperative principle in "efficiency", but I think its implications are much deeper and more related to wordplay -- if nothing is ambiguous, pragmatics basically ceases to exist. Regarding the necessity of garden path sentences, I merely meant that I didn't think it particularly advantageous to have garden path sentences in a language. Feb 12, 2018 at 2:26