This article needs attention from an expert in linguistics. Please add a reason or a talk parameter to this template to explain the issue with the article. WikiProject Linguistics may be able to help recruit an expert.(July 2016)
Some of the issues below are commonly recognized as unsolved problems; i.e. it is generally agreed that no solution is known. Others may be described as controversies; i.e. although there is no common agreement about the answer, there are established schools of thought that believe they have a correct answer.
Is it possible to formally delineate languages from each other? That is to say, is it possible to use linguistic (rather than social) criteria to draw a clear boundary between two closely related languages with a dialect continuum between their respective standard forms (e.g. Occitan and Catalan)?
Unclassified languages (languages whose genetic affiliation has not been established, mostly due to lack of reliable data) make up about 38 of the 6,000–7,000 languages spoken in the world. An additional 45 languages are classified as language isolates, with no demonstrable relationship to other languages.
Can we use the comparative method to reconstruct back to an arbitrary time depth, or do we need new methods to reconstruct the distant past of languages? Is there a time depth beyond which we cannot reconstruct? Can we ever demonstrate that all languages are ultimately related to each other, or that they aren't?
Is the human ability to use syntax based on innate mental structures or is syntactic speech the function of intelligence and interaction with other humans? The question is closely related to those of language emergence and acquisition.
The language acquisition device: How localized is language in the brain? Is there a particular area in the brain responsible for the development of language abilities or is it only partially localized?
What role does linguistic intuition play, how is it formed and how does it function? Is it closely linked to exposure to a unique set of different experiences and their contexts throughout one's personal life?
Is perfect computational word-sense disambiguation attainable by using software? If yes, how and why? If no, why? (This presupposes the solution to the unsolved problems in the other areas of linguistics as a basis.)
To what extent are termbases reliable and can their reliability be measured objectively? If yes, how and why? If no, why? What is the relationship between termbases and individual subjectivity and can subjectivity about word sense disambiguation be overcome at all or is it a natural result of different experiences in one's unique personal life?