[gabemart]

I found this article fascinating and satisfying.

I'm curious about the desire to reduce ambiguity, which seemed to be emphasized as a motivation for the creation of Ithkuil and some of the other languages mentioned.

Is it desirable to completely eliminate ambiguity? I can see why it would be desirable in a scientific paper or a public political debate. But in everyday interactions, (intentional) ambiguity plays many important roles.

In my experience, politeness is bolstered by some level of ambiguity. Rather than explicitly state your needs, desires or opinions, you imply them at some level of abstraction, allowing other participants in the conversation to accept or decline more easily. Imagine Jessica who has brought two friends who don't know each other to see a play. They chit-chat a little afterwards, then Jessica goes home early leaving two virtual strangers to have a drink together. It's not hard to imagine the conversation going like this:

A: "Did you enjoy the play?"

B: "It was very interesting. I thought the stage dressing was a little unconventional."

A: "Yes, I noticed that too. Very creative. I was intrigued by the style of the narration. It really let the audience write the story for themselves."

B: "It certainly didn't constrain the imagination did it? I couldn't help noticing that many of the actors took a somewhat avant-garde interpretation of the source material."

A: "Yes, as if they didn't want it to seem like they were 'acting', so to speak?"

B: It was awful wasn't it!?

A: Thank god! Yes, worst thing I've ever seen!

Ambiguity allows subtle social cues (not so subtle in my example!) that avoid direct confrontation when it might be uncomfortable. If one person loved the play and the other hated it, they each might want to avoid offending the other.

Intentional ambiguity plays an important role in other social interactions like dating or friendship-making. Correct use of ambiguity protects feelings, demonstrates subtlety and good judgement, and avoids non-productive conflict.

In artistic expression too, ambiguity is often intentional or even necessary to the effectiveness of the work. Consider a poem like "My Papa's Waltz" [1]. Does it describe happy memories of the narrator's father, or dark memories of childhood abuse [2]? Can it describe both? Is there something in between? The ambiguity isn't a byproduct of imprecise language. The ambiguity is the meaning. To resolve it is to remove the point of the work. The poem cannot be effectively communicated in any medium that does not allow for the existence of ambiguity.

[1] http://www.poetryfoundation.org/poem/172103

[2] 'Yet, this poem has an intriguing ambiguity that elicits startlingly different interpretations. Kennedy calls it a scene of "comedy" and "persistent love", and Balakian, in part, labels it a "comic romp" (62). In contrast, Ciardi sees it as a "poem of terror"' - from http://www.mrbauld.com/exrthkwtz.html

[jasode]

>, politeness is bolstered by some level of ambiguity. Rather than explicitly state your needs, desires or opinions, you imply them at some level of abstraction, allowing other participants in the conversation to accept or decline more easily.

Steven Pinker explores this point in an entertaining presentation[1] for RSA. He also covers other language topics such as spacetime encoding, and profanity but the last part analyzes the need for ambiguity in a language. I deep-linked into the relevant portion of the presentation although the the entire talk is very enlightening. It's worth rewinding to the beginning to watch the entire talk. The 2nd youtube video[2] is the mostly the same material but it's the older one he presented at Google TechTalks.

[1]http://www.youtube.com/watch?v=5S1d3cNge24#t=32m55s

[2]http://www.youtube.com/watch?v=hBpetDxIEMU#t=40m38s

[gsnedders]

To quote Levinson (2000): "inference is cheap, articulation expensive, and thus the design requirements are for a system that maximizes inference". All natural languages [citation needed] rely heavily on inference — simply because it allows one to minimize what's said. That's not to say that everything maximally relies on inference — politeness phenomena clearly demonstrate otherwise, often being "needlessly" verbose.

[andreasvc]

I agree, and additionally I would say we shouldn't remove ambiguity for a more basic reason: humans are very good at resolving it. There may be lots of other areas where we have systematic biases such as estimating probabilities, but dealing with ambiguity is actually one of our big strengths so it doesn't make sense to spend so much energy on avoiding it.

This is a bit speculative, but I have the feeling there is a certain kind of personality that is typically fascinated with this idea of being able to rule out ambiguity. It seems to me a tendency to overextend the idea of mathematical rigor to other areas of life.

[ar-jan]

The point has also been made that ambiguity is a functional property of language, because it is more efficient from an information-theoretic point of view: context also provides information, but a purely unambiguous language would have to express that information anyway. Ambiguity also allows reusing words and sounds. (http://www.sciencedirect.com/science/article/pii/S0010027711...).

[igravious]

Couldn't agree more. Seven Types of Ambiguity published as far back as 1930 by William Empson launched a school of criticism called New Criticism. A definition of ambiguity is then "alternative views might be taken without sheer misreading." For Empson poetry is heavily reliant on ambiguity. And, arguably, poetry is language at its most wrought with ideas most distilled.

[moomin]

One of the threads in David Brin's Jijo trilogy is that development in the galaxy is being held back by (amongst other things) designed, unambiguous languages.

[JoeAltmaier]

Ambiguity is necessary for art. If there is no room for interpretation, then its just a photo-of-words.

[ginko]

I also wonder if a concise language like this wouldn't make lying easier since you can manipulate the meaning of your words by so slightly altering their spelling and pronounciation.

[alxndr]

I think an unambiguous human-speakable language would be a great candidate to teach to a computer.