The Aikanã and Kwaza languages are genetic isolates, which means that they are apparently not related to any of the world’s languages or linguistic families. There are notable similarities between the Aikanã and Kwaza languages that are also shared with Kanoê (the other isolate of southeastern Rondônia). Although there is no compelling evidence that they are genetically related to one another, the possibility of a long-distance genetic relationship cannot be excluded. At any rate, their lexicons are almost completely different, and those etymological correspondences that can be found are usually so close that they must be the result of borrowing. Areal diffusion may explain part of the similarities between these languages since these language groups are known to have married and forged alliances across ethno-linguistic lines.
Aikanã is spoken by around 170 individuals. The majority of its speakers live in the indigenous reserve Terra Indígena Tubarão-Latundê. While children born to parents who are both Aikanã tend to grow up learning the language at home, a significant number of interethnic marriages, especially between the Aikanã and members of the local Nambiwkaran groups or with non-indigenous Brazilians, have led to a number of households adopting Portuguese as the language of daily communication.
Kwaza is spoken by around 25 individuals belonging to three families. One family, of which only the oldest generation speaks the language, lives in the indigenous reserve Terra Indígena Tubarão-Latundê, another family lives mainly in the nearby town of Chupinguaia, and a third family of Kwaza speakers lives in the Terra Indígena Kwazá do Rio São Pedro. The two latter families are related, and the language is still acquired by the youngest generation.
Both Aikanã and Kwaza have contrastive oral and nasal vowels. Aikanã has a voiced interdental fricative consonant /ð/ that can be nasalised and a close front rounded vowel /y/. Kwaza has labial and apico-alveolar implosive consonants /ɓ/ and /ɗ/ and relatively many vowel contrasts.
In terms of morphosyntax, both Aikanã and Kwaza are morphologically rather complex, especially with regard to the verbs. They are predominantly suffixing languages and are characterised by a certain degree of polysynthesis. Both languages have many classifying and valency changing suffixes. Both languages have a dependent-marking pattern in possessive expressions. Many of these morphosyntactic traits are shared with Kanoê and some with other languages of the region. Some traits are shared only between Aikanã and Kwaza, such as the anticipatory switch reference marking system that indicates whether the subject of the next clause will be different or not from that of the current clause. Both have an alternative head-marking possessive construction that concerns only the third person and involves a somewhat similar morpheme (Aikanã –deri, Kwaza –tjate).
The person marking systems of Aikanã and Kwaza are very different, however. Whereas Kwaza displays an inclusive/exclusive distinction, Aikanã does not. Whereas Kwaza has one single paradigm of person inflexions for all verbs, Aikanã has about twelve different verbal inflexion classes with different sets of person markers. Whereas Kwaza person marking is exclusively suffixing, several Aikanã verb classes require prefixation of person markers. For linguists, language isolates are particularly interesting because they may have properties that are very rare or even unique among the world’s languages. So far, Kwaza is the only attested language in which reduplication is determined by morphological boundaries as opposed to phonotactic boundaries –in this case reduplication of bound person markers with different syllable structures. Whereas Kwaza has been described relatively thoroughly, Aikanã is still awaiting an exhaustive description and analysis.
References:
Bacelar, Laércio N. 2004. Gramática da língua Kanoê. PhD dissertation, Katholieke Universiteit Nijmegen. [http://webdoc.ubn.kun.nl/mono/b/bacelar_l/gramdalik.pdf].
Crevels, Mily & Hein van der Voort. 2008. “The Guaporé-Mamoré region as a linguistic area.” Pieter Muysken (ed.) From Linguistic Areas to Areal Linguistics, Studies in Language Companion Series 90, Amsterdam/Philadelphia: John Benjamins, p. 151-179.
van der Voort, Hein. 2003. Reduplication of person markers in Kwaza, Acta Linguistica Hafniensia, 35, 65-94.
———. 2005. Kwaza in a comparative perspective. International Journal of American Linguistics 71(4): 365–412.
———. 2009. Possessive expressions in the Southwestern Amazon. The expression of possession. William B. McGregor (ed.), 343-388. Berlin / New York: Mouton de Gruyter.
———. 2009. Reduplication and repetition of person markers in Guaporé isolates. Morphology, 19/2: 263-286.
http://www.springerlink.com/content/1545301353748536/fulltext.pdf