Ikpeng is a member of the Cariban language family spoken by approximately 584 individuals in the Xingu Indigenous Park (PIX, for Portuguese ‘Parque Indígena do Xingu’), in the northeast region of the state of Mato Grosso of central Brazil.1 Although the Ikpeng, along with the other ethnic groups of the PIX, have been in constant interaction with the national society since the middle of the 20th century, language vitality is strong and all ethnic members of the group use Ikpeng as their day-to-day mode of communication. Figure 1 depicts the current location of the main Ikpeng village, Moygu, adjacent to the Pavuru Indigenous Outpost.

According to Chagas (2017, 2022), the Ikpeng currently reside in six different villages, with 57% of the total Ikpeng population inhabiting the village of Moygu. The other villages, in decreasing order of permanent population, are Tupara (17%), Arayo (12%), Rawo (6%), Paranoa (5%), and Kurure (3%). Women account for 54% of the Ikpeng, and nearly 80% of the Ikpeng are age 35 or younger. This fits the typical demographic profile for Indigenous groups in Brazil.

The Ikpeng language is spoken by all members of the group and is the language of choice for domestic and day-to-day interactions in Ikpeng villages (Chagas 2017, 2022). Intergenerational transmission is strong, with all children learning Ikpeng as their first language. Nevertheless, contact with the national language, Portuguese, starts early for the younger generations, while older individuals have less acquaintance with the language. Women have a weaker command of the secondary language than do men. Portuguese is used in interactions with the encroaching non-Indigenous society, and inter-ethnic marriages involving other Indigenous groups of the PIX have also brought many of these languages to Ikpeng villages. So there are Ikpeng people who claim some command of Kayabí (Tupi-Guarani), Waurá (Arawakan), Kamayurá (Tupi-Guarani), Mehinaku (Arawakan), Kalapalo (Cariban), and the isolate Trumai.

Linguistic variants associated with the speech of older generations have prestige and are usually seen as the ‘true Ikpeng’ by younger individuals, in line with the generalized respect shown to elders in Ikpeng society. All Ikpeng villages are served by formal schools from elementary through high school following Brazilian legislation that demands that Indigenous schools should be sensitive to the specific reality of the Indigenous people that attend them. Although the teachers are all native Ikpeng speakers, some of the more technical subjects such as biology or mathematics require the use of Portuguese as a language of instruction, in part due to a lack of pedagogical resources written in Ikpeng. The main hindrance, however, remains the limited elaboration of codified technical vocabulary and metalanguage in Ikpeng for the relevant subjects, which is a condition for adequate pedagogical materials to be produced in the first place. The same asymmetry in the availability of written material likely accounts for the fact that more than half of Ikpeng children report greater proficiency reading and writing in Portuguese than in Ikpeng (Chagas 2017). While they recognize that knowledge of Portuguese is useful, most Ikpeng claim that formal education in and about the Ikpeng language is also important to them.

The first Western reports of the Ikpeng, then known as ‘Txicão’, date to 1944, thus predating the creation of the PIX. They mention the group as attackers and enemies of Indigenous groups located in Upper Xingu environs, in particular Waurá and Mehinaku communities along the Curisevo and Batovi affluents of the Xingu river (Galvão & Simões 1965; Menget 2001: 80; again see Figure 1). Attacks attributed to the Ikpeng would target these and other Upper Xingu groups, continuing until 1960. According to Simões (1963: 101), comparison of Ikpeng material culture with earlier reports by travelers and explorers allows for a conjectural identification between the Ikpeng and the ‘Kabischi’ of Meyer (1900), the ‘Apiaká’ of Koch-Grünberg (1902) and the group identified by the Suyá people as ‘Cuiaáus’ to Steinen (1885).

Anthropologist Patrick Menget (1977) tried to reconstruct aspects of Ikpeng pre-history based on memories and recollections of Ikpeng elders with whom he conversed during the 1960s. Menget (2001: 74) identifies the Ikpeng as living peacefully with the Tupian-speaking Xipaya along the Iriri river around 1850. The first peaceful interaction between the Ikpeng and members of the national society were established only in 1964, and in October of the same year the first photographic registers of the Ikpeng, as well as collection of Ikpeng vocabulary, were made (Galvão & Simões 1965: 6–7). At this time the Ikpeng were living close to the Jatobá river, an affluent of the Ronuro River (again see Figure 1) and were badly affected by disease and malnutrition, likely resulting from contact with encroaching ranchers, miners, and prospectors. This is the background against which the Ikpeng agreed to be transferred to the PIX in 1967. Five days of navigation downriver in a ferry were followed by a complex process of negotiation involving representatives from different ethnic groups, since the Ikpeng were not originally part of the Upper Xingu cultural zone. Indeed, the Ikpeng were long-time enemies and adversaries of some of the Xinguano groups who lived there, in particular the Waurá and the Mehinaku. (See Menezes 2001 for a fascinating history of the political, economic and legal context of the creation of the PIX.) The Ikpeng were re-located to the middle course of the Xingu River in 1980. The Pavuru Indigenous post, along with the main Ikpeng village, Moygu, were founded in 1985 (Menget & Troncarelli 2003).

While there have been a number of studies of Ikpeng culture, sociality, and demography (see, e.g., Galvão & Simões 1965; Menget 1977; Rodgers 2002; Simões 1963; and Taffarell & Januário 2010), systematic investigation of the language is more recent. Charlotte Emmerich’s (1972) master’s thesis, published in 1980, was the first scientific investigation of the language. Emmerich presents an essentially phonemic description of Ikpeng segmental phonology and morphophonology; Emmerich (1991, 1994) discuss other aspects of Ikpeng phonology. Cilene Campetela’s (1997) master’s thesis describes Ikpeng phonology and the morphosyntax of independent clauses, and her (2002) doctoral dissertation provides an in-depth description of aspects of the language’s prosody and intonational phonology, focusing on their interactions with allomorphy and word-formation processes. Frantomé Pacheco’s (1997) analysis of Ikpeng phonology differs in some ways from that of Emmerich (1972), in addition to offering a preliminary description of certain aspects of the language’s grammar. Pacheco’s (2001) doctoral dissertation focuses on the morphosyntax of the Ikpeng verb, including valence-changing processes, nominalization, and dependent verb forms. Pacheco has been the most prolific author writing on Ikpeng thus far, with 12 published papers and three book chapters on the language (in addition to the above, see Pacheco 1998, 2003, 2005, 2006, 2009).

Angela Chagas’s doctoral dissertation is a lexico-semantic study of the morphosyntax of Ikpeng verbs. Chagas also headed a documentation project on Ikpeng (2009–2012) within the purview of the Projeto de Documentação de Línguas Indígenas (PRODOCLIN), an initiative of the Museu do Índio/FUNAI in partnership with UNESCO.2 The PRODOCLIN project yielded 33 hours of audio-visual documentation of traditional activities including fishing, housebuilding, festivities, monologues, and traditional narratives. These archives are now held by the Museu do Índio, the entity responsible for a publicly accessible repository on the endangered languages of Brazil. Unfortunately, federal investment on all matters related to Indigenous peoples has been drastically reduced since 2016. Still, 22 of the recorded narratives were published in a monolingual storybook, Wonkinom Mïran (Chagas & Costa 2014).

The typological profile of Ikpeng is typical for a Cariban language (Derbyshire 1999; Gildea 2012). Unlike other languages of the family, Ikpeng lacks a separate reflex of Proto-Cariban *ə (see Meira, Gildea & Hoff 2010), and neither length nor nasalization are contrastive. Clusters of up to two consonants are common in word-medial position, likely a result of the pan-Cariban process of syllable reduction (see Gildea 1995), which gave rise to a relatively complex morphophonology that is supplemented by other processes such as vowel harmony, vocalic ablaut, and both root- and affix-allomorphy. Ikpeng is predominantly head-marking and head-final, with most of the morphology showing up in verbs; nonetheless, a complex system of postpositions is found as well as mechanisms for marking oblique NPs and certain semantic roles. Ikpeng morphology is primarily but not exclusively suffixing; verbs have valence-changing prefixes (reflexive) and suffixes (causative) as well as TAM suffixes.3 As elsewhere in the family, transitivity is an important parameter. Transitive verbs feature a single morphological slot for argument-indexing, with a person hierarchy determining which argument, A or P, gets morphologically coded in the verb. Intransitive verbs show a lexically determined split-S pattern. As in other Cariban languages, Ikpeng has a plethora of mechanisms for nominalizing verb bases, many of which function then as heads of subordinate or dependent clauses. There are several suffixes that form denominal verbs (Chagas 2013), although many of these seem etymological only and hard to discern on strictly synchronic grounds.

The authors are currently engaged in the documentation of Ikpeng with the goal of better describing the language’s morphophonology, phonological patterning, and the grammar of possessive constructions and dependent clauses. We also aim to contribute to the education of Ikpeng teachers by producing improved materials for use in Indigenous schools. The project, directed by Angela Chagas, is funded by the Brazilian National Council of Research (CNPq project number: 430958/2018-8). Based on fieldwork at the Pavuru Indigenous Post in the Xingu Indigenous Park, the authors have produced recorded narratives and a bank of elicited data to support phonological and grammatical analysis. These materials will be archived at the Museu Paraense Emilio Goeldi, Pará, Brazil. At this point the data is at an advanced stage of transcription, and a series of papers on the description of Ikpeng phonology, morphophonology, and aspects of its morphosyntax are in preparation (e.g., Carvalho, Chagas & Vasconcelos in press).