Kawahíva is a severely endangered and understudied Tupí-Guaraní (TG) language of Western Brazil. Members of the Kawahíva community use this name to refer to themselves as a community and to refer to someone as an Indigenous person.

There are roughly 560 Kawahíva speakers out of a current population of 1070 individuals across eight ethnic communities, with each community representing a different ethnic dialect (Sampaio 1997, 2001). In almost all the communities, the language is severely endangered. For example, only three of the 12 Júma people speak the language; only one of the 50 Jiahui people is a speaker (Moore, Galucio & Gabas 2008). Figure 1 presents the rough location of each group’s village. The Jupaú, Parintintin, and Tenharin are distributed over several villages, while the other communities each reside in one village.2

Other Kawahíva dialects previously noted in the literature are now considered extinct, including those spoken by groups that once lived near the Machado River and its tributaries, including the Paranawat (Nimuendajú 1981), Wiraféd (Nimuendajú 1955, 1981), Takwatip (Nimuendajú 1948, 1981; Lévi-Strauss 1955: 379–439), and Ipotewát (Lévi-Strauss 1955: 379–439). Lévi-Strauss also mentions the use of the language by people who were already almost extinct at the time, including the Tucumanfét and the Jabotiféd, who lived near the Machado/Ji-Paraná River and the Mialat who inhabited the Leitão River region. The last known speaker of the Capivari dialect, Pitanga, passed away in the fall of 2022 (Hanmin Kin, personal communication, 2023-06).

Previous works suggest a language-internal division of Kawahíva into southern and northern dialects (Aguilar 2015; Marçoli 2018), but no evidence from shared innovations has been brought forward to support this division. The closest sister languages to Kawahíva in the Tupí-Guaraní family are Apiaká and Kayabí (Aguilar 2015; Michael et al. 2015; Rodrigues & Cabral 2002).

According to oral histories, the Kawahíva originated from a common ancestor but later split into different groups due to disagreements.3 An ethnohistorical map by Nimuendajú (1981) also supports this theory, adding that the descendants of this group once inhabited the region near the mouth of the Tapajós River and fled the area because of their traditional enemies, the Munduruku. More details about the Kawahíva migrations can be found in Menéndez (1989).

Kawahíva territories are located in linguistically diverse areas, including the state of Rondônia (Galucio et al. 2018). The Kawahíva strongly resist outside linguistic influences. The likeliest language-external influence is one lexical calque from Brazilian Portuguese found in Júma and Jupaú: the word for ‘person’, ahe, is also used as a pronoun, much as gente ‘person’ is used in modern Brazilian Portuguese. (In Kawahíva’s sister languages, the cognate form is used only as a 3rd person pronoun.) A verb-initial syntactic pattern in the Júma and Jupaú dialects is also suspicious due to the proximity to Wari’, which shows a rigid VOS pattern (Joshua Birchall, personal communication, 2021-08). Finally, it has been suggested that Tenharin is the source for Pirahã pronouns (Thomason & Everett 2010), implying contact between these communities.

Data from a language survey (Moore et al. 2008) and the author’s fieldwork show that almost all Kawahíva varieties are severely endangered. This is evident from the difference between the population and the number of speakers. Apart from the Tenharin Marmelos community, Kawahíva children no longer learn Kawahíva as their first language. Children are becoming passive bilinguals in some communities like Amondawa, Júma, and Jupaú. There has been a rapid loss of many dialects within a single generation.

Systemic discrimination and prejudice brought by outsiders have undermined the value and recognition of Kawahíva language and culture. This is evident even at basic levels of language rights, such as the right to have a Kawahíva name. Among the Jupaú, the generation that was born after contact in the 1980s received both a Kawahíva name and a non-Indigenous name, the latter being used both within and outside the villages. However, the younger generation of Kawahíva received only a non-Indigenous name. This is likely due to uncomfortable experiences people had using their Indigenous names, such as having their names mispronounced in public, with the mispronunciations often matching Portuguese words. For example, the male name Puré [puɾɛ] has been mispronounced as purê [puɾe] ‘mashed potato’ in Portuguese. Others have had their names misspelled on their ID documents, which then became the actual pronunciation. Such is the case of Mba’yta [mbaʔɨta], which was changed to Maytá [maita]. Contact with outsiders has also introduced detrimental practices such as alcohol consumption and smoking (Kracke 1973; Peggion 2005:194).

Despite Kawahíva being the official language for elementary school instruction, the lack of materials, inadequate teacher training, and insufficient institutional support have hindered effective use of the language in schools. Opportunities to learn the language at elementary levels are already limited, and they become null once students move to non-Indigenous schools for higher levels of education. Only in Tenharin Marmelos students are taught in the Indigenous language across all subjects, providing a unique opportunity to preserve the language. According to conversations with schoolteachers and young adults who studied there, this approach has been successful.

Linguistic knowledge of the language is mostly restricted to extant varieties thanks to SIL missionaries who produced wordlists, phonemic analyses, and sketch grammar (Betts 1981; Pease 2007; Pease & Betts 1971). The most substantive work done by the missionaries has been on the lexicon of Parintintin and Tenharin, with a few publications on grammatical aspects of the language. This includes a phonological overview of Parintintin with remarks on nasalization phenomena (Pease & Betts 1971), a 78-page Parintintin grammar (Pease 2007), a Parintintin-Portuguese dictionary (Betts 1981), and a posthumously expanded Kawahíva-English dictionary with additional data from Amondawa, Jupaú, and Tenharin (Betts 2012). Works on other varieties, including word lists and brief grammatical descriptions, are listed on the Language and Culture Archives page of the Summer of Institute of Linguistics but are not available online.

Non-missionary linguists have expanded the language description to other varieties (e.g., Amondawa, Jupaú, Júma) and their grammar. Topics addressed include classification of the language group (Sampaio 1997, 2001), a phonological comparison (Marçoli 2018), and studies of verbal agreement (Martins & Vezzaro 2017; Vezzaro 2015). The author’s work has focused on nasal harmony, reduplication, reported speech, verb-initial clauses, realis mood, the morphosyntax of relative clauses and noun phrases, and verbal agreement and unaccusativity (Dos Santos 2019a, 2022, 2023).

Ongoing language documentation with Kawahíva communities is described in Dos Santos (2019a). These efforts have led to a deposit in the California Language Archive (Dos Santos 2019b) and a multimedia dictionary.4 A Kawahíva-Portuguese text collection and a sketch grammar are currently in preparation. The language documentation collection contains recordings of elicitation sessions, traditional and personal stories, songs, and photographs. This is the first documentation, as opposed to a description, of the language, as none of the missionaries’ data was recorded. Currently, there are over 13 hours of Júma texts collected, of which 12 hours have been fully transcribed and translated. Over 16 hours of Jupaú texts have been collected, of which three hours are transcribed and translated. Additionally, a partnership between Júma, Jupaú, UNESCO, and Museu do Índio/FUNAI resulted in a Kawahíva keyboard and dictionary app.5 The dictionary is based on the Júma dialect and contains over 1500 entries. It also includes audio and photographs.

Revitalization efforts have focused on addressing social injustices to prevent continued language loss, including developing a practical orthography and organizing a literacy workshop and panel on a university campus in Humaitá, Amazonas, in June 2023. The panel, part of which is depicted in Figure 2, aimed to highlight how oppression, stigmatization, marginalization, and other factors contribute to the endangerment of Kawahíva.