Language Misclassification

Length: 750 words

Purpose:

• To show that language misclassification is a present yet commonly ignored threat for historical bookkeeping
• Many languages have a written alphabet, but some do not. Those that are unwritten and are often spoken by families with little generational wealth are in the greatest danger of being forgotten.

Important Points to Address:

• Who is in charge of classifying languages, and why do classifications sometimes differ?
  • Look at Ethnologue and Glottolog
  • See next point
• The Bazigar language can be used as a case study for the article
  • It is a South Asian language, classified as Dravidian by Ethnologue and “Unclassifiable” by Glottolog
  • Still, when spoken, it is extremely difficult to differentiate from the Punjabi language
    • It is very difficult to understand why these language classification agencies do not give it a designation of being a Punjabi sub-dialect because the agencies are comprised of dozens of linguists across countries who each have individually differing reasons for language classification
    • However, the classification itself is still incorrect
• How does this affect future generations who may wish to learn the language? This is the main point of the argument so put a great emphasis on this section
  • Because those who often speak unwritten languages come from a background where they do not have the opportunity to digitize their language, the responsibility lies in the hands of linguists who classify such languages…
    • Such that those who wish to learn it have something to compare it to even if it does not give a 1:1 translation
  • If the classification is incorrect, then the future translation will also be incorrect (due to a limited digital footprint of the language)
• Possible Solution
  • Most issues with language classification stem from the fact that there is no centralized translation accuracy tool (no way to compare the similarity between words, spellings, phonetics, etc. between different languages) so linguists’ and agencies’ classifications are prone to human error
  • Ontology systems could be used in combination with artificial intelligence to map out the relations and proximity between languages, using a predetermined phonetic alphabet*
    • * If we first wrote the words based on how they sound in Phonetic English or some other agreed upon, possibly a new, phonetic alphabet
    • * This could allow better classification of even unwritten languages