Xhosa

From Apertium
Revision as of 20:36, 6 January 2018 by SlingKing7 (talk | contribs)
Jump to navigation Jump to search

The Xhosa Language

Xhosa (Wikipedia: Xhosa language) is a (Wikipedia: Nguni Bantu) language spoken mainly in Africa. Its widespread use is not very common and only has a small number of individuals enacting the language (11 million)

  • Zulu and Xhosa are relatably the most similar due to them both being Nguni transcripts (predominant in Southern Africa)
  • Xhosa is also similar to the languages of Swati and Ndebele which are spoken among the Bantu people


Apertium Language Pairs

Currently, on Apertium, the language of Xhosa is recorded to have one language pair:

(list of language pairs)

Computational Linguistics

Xhosa Cross Linguistics

The Tone Analysis of Xhosa

Comparative Studies

Xhosa Literary + Academic Study (Rhodes University)

Comparative Study: Zulu - Xhosa


Linguistic Grammar

Xhosa contains multiple prefixes and suffixes which are attached to root words. Thus, the language is declassified into fifteen morphological classes or genders. Furthermore, the language is unique based on its tones, the phonemic low, and high tones;

  • they are a [à], á [á], â [áà], ä [àá]. Long vowels are phonemic but are usually not written, except for â and ä

The usage of uncommon consonants is dominant throughout the language in the version of clicks. The language uses 21 clicks (7 dental), however, the number of clicks varies based on each region (Namibia and Botswana primarily)


Xhosa Pronunciation + Language Set: "Omniglot Writing Systems and Languages", a full context of the usage of grammar within the Xhosa language



Bilingual/Monolingual Dictionaries

(most are mainly decoded within English subtexts and contexts)

Elaboration on Machine Translation:

  • Machine translation is a sub-field of computational linguistics that investigates the use of software to translate text or speech from one language to another.
  • The isiXhosa language is a low resource language for which there are not many morphological analyzers, so much of the research is compiled into putting together a parallel text for which a translation model can be trained

Monolingual/Parallel Corpora