Can a language be completely non-ambiguous?
All human languages are ambiguous at different levels: lexical and semantic. For example:
Lexical ambiguity (homonyms, homophones, homographs, etc.):
– How to pronounce ea in great, dead and meat, or sas in Kansas and Arkansas?
– Which lead are we talking about, lead (the metal pronounced led) or lead (the verb pronounced leed in the present tense and led in the past tense)?
– In French, the same sound can be spelled sans, sang, s’en, c’en, sent or sens.
– Most words have several meanings, such as note or lie.
Semantic ambiguity:
– Who is nude when I say “I like to paint her nude“? Hmm, maybe both?
– Does “he’s left” mean that he’s gone or that he stayed behind?
– In French, un hôte can be either a guest or a host.
An interesting initiative is Lojban, “a carefully constructed spoken language. It has been built for over 50 years by dozens of workers and hundreds of supporters” with, among others, the objective of removing most if not all ambiguity.
From their website:
Lojban is an experiment in language — the grammar is regular, simpler than most natural languages, but complex in its own unique way.
Lojban allows the expression of emotion, using words called attitudinals, which are essentially spoken emoticons, expressing your feelings.
Lojban allows you to communicate concisely without unnecessary or undesired details. For example, you don’t have to always think of what tense (past, present or future) to use in a verb when it’s already clear from context.
Lojban has 1341 root words, covering a broad section of human experience. These words may be combined into compounds, allowing precision in meaning. Futhermore, Lojban makes provisions for borrowings, integrating them into the Lojban morphophonologic system.
Lojban is syntactically unambiguous.
Lojban is machine parsable, allowing potential new explorations in the fields of artificial intelligence communication and machine translation.
Lojban is culturally neutral, as far as such things are achievable. The base vocabulary was generated using an algorithm and words from the (at the time of its creation) 6 most spoken languages on Earth, averaging them out.
Lojban has a flexible pronunciation. Lojban phonemes (basic units of sound) allow many realizations, for example, the r may be pronounced as an English, French, or Spanish r.
We are not planning to start offering translations in Lojban anytime soon as this is an experimental language mostly used by scientists and researchers, but the effort of this community is worth mentioning!
Computational linguists have tried for many decades to fight for “computational disambiguation” in English and other languages. A non-ambiguous language could be perfectly understood by a machine, as it theoretically would be context-dependent. Among others, machine translation would work perfectly if this problem could be solved, or if we would write in lojban!
This said, I believe that a certain level of ambiguity must be treasured. Comedians, poets, novelists or humorists use it every day. Ambiguity is the essence of metaphors, allegories, paradoxes, connotations, word plays, etc. Would you like to live in a world and speak a language that would make all of these impossible?