Context
Upper Necaxa is a member of the Totonac-Tepehua family of languages, an isolate linguistic group of Northern Puebla State, Mexico, and adjacent areas of Veracruz and Hidalgo. It is currently spoken by around 3,400 people, most aged 40 or older, in four communities—Patla, Chicontla, Cacahuatlán, and San Pedro Tlaloantongo—in the Necaxa River Valley in northeastern Puebla State. All but the oldest speakers are bilingual, and in most households the language of child-rearing has been switched to Spanish by parents who see proficiency in the dominant language as an advantage for their children. At least one other Totonacan language (Misantla Totonac) has only a handful of speakers remaining (MacKay 1999) while another (Ozumatlán) is spoken by fewer than 2,000 people (Gordon 2005). Most of the indigenous languages of Mexico, like the majority of the world’s languages, are in grave danger of extinction, with potentially tragic results for scientists and speakers of these languages alike (Krauss 1992).
The Totonacan branch of the Totonac-Tepehua family has traditionally been analyzed as consisting of three or four mutually unintelligible languages—Misantla, Papantla, and Sierra (Arana 1953; García Rojas 1978), the last group sometimes being subdivided into Northern and Southern Sierra. Recent work indicates that the separation within some of these varieties is great enough to prevent naïve mutual intelligibility, the usual criterion for the classification of speech varieties as separate languages. The distinctions between Upper Necaxa and other members of the Northern Sierra group, within which it has been classified, seem particularly marked and include a number of distinctive lexical, grammatical and phonological changes (Gordon 2005; Beck 2000a, 2006c). In some cases, the divergent features of UNT seem to link it more to Southern Sierra and Papantla, suggesting that it may be in some sense an inter-mediate variety crucial for accurate historical reconstruction, a project currently being undertaken in collaboration with colleagues in Computing Sciences at the University of Alberta (Kondrak, Beck & Dilts 2007). As a result of the PI’s research to date, UNT has been recognized by the Mexican government through INALI (Instituto Nacional de Lenguas Indígenas), and has recently been assigned an ISO language code (TKU), as has a neighboring Totonacan language, Tecpatlán Totonac (ISO TCW).
Until fairly recently, the Totonac-Tepehua family has been largely undescribed. The earliest work is a grammar written by a Spanish priest (Bonilla 1752). A few languages of the family have been studied by evangelical missionaries from the Summer Institute of Linguistics (SIL), who have produced three moderate-sized dictionaries (Aschmann 1962, 1973; Reid & Bishop 1974) and some articles on specific topics (e.g. Aschmann 1946a, 1946b, 1952-53, 1963; Aschmann & Wonderly 1952; Bishop 1979). The first grammatical work on Totonac by an academically-trained linguist is a study of a variety of Sierra Totonac (Coatepec), written in the 1940s and re-printed in Spanish as McQuown (1990), while aspects of Tlachichilco Tepehua grammar are dealt with in Watters (1985, 1987, 1988). Northern Totonac has received no previous attention from linguists, although the Apapantilla variety has been documented to some extent by members of the SIL (Reid et al. 1968; Reid 1979, 1991; Bishop & Reid 1979).
More recently, there has been an increased interest in the language family among academics. New stud-ies are currently underway of Pisa Flores Tepehua and of the Sierra Totonac varieties spoken in Zapotitlán and Filomeno Mata. Misantla Totonac is described by MacKay (1991, 1994, 1999) and Papantla Totonac has been written on by Levy (1987, 1990, 1992, 1994, 1996, 1997, 1999a, 1999b, 2002a, 2002b, 2002c, 2003, 2004), while Huehuetla Tepehua is the subject of a recent PhD dissertation (Smythe Kung 2007). An MA thesis from CIESAS (Centro de Investigaciones y Estudios Superiores en Antropología Social) examined the acquisition of Olinta Totonac (Tino Antonio 2006), and another CIESAS student is working on the Totonac of Ozelonecaxtla. May 2007 saw the first ever International Conference on Totonac-Tepehua Languages, with 11 participants from Canada, the United States, and Mexico, including three graduate students and three native speakers of different Totonacan languages.
Previously undescribed, UNT is now beginning to surface in the academic literature (Beck 1999a, 1999b, 2000a, 2000b, 2001a, 2001b, 2001c, 2002, 2003a, 2003b, 2004, 2006a, 2006b, 2006c, 2007a, 2007b, to appear a, to appear b, to appear c, in progress). As a result of previous SSHRC-sponsored projects, the PI currently has a lexical database of more than 8,900 entries and has recorded some 19,676 word-tokens and example sentences, as well as 12 hours of oral texts (folk tales, personal narratives, and interviews), 3 hours of which have yet to be transcribed and analyzed. This research has shown UNT to be a typologically unusual language with a number of structural features that set it apart even from its sister languages. Phonologically, the language makes distinctions in vowels for both length and laryngealization, and has an unusual consonantal inventory including three ejective fricatives (/s’/, /ʃ’/, and /ɬ’/) but no ejective stops (Beck 2006c). Morphologically, UNT is a polysynthetic agglutinating lan-guage whose verbs combine 10 prefixal positions with 14 suffixal positions (Beck, Holden & Varela n.d.), many affixes showing variable rather than templatic ordering (Beck to appear c). Others occur in conventionalized combinations with non-transparent meanings — essentially, morphological idioms (Beck 2007a). The language also makes use of bodypart prefixation in locative and other constructions (Beck 2004), and compounds active verbs with stative posture verbs to create complex aspectual expres-sions that go well beyond the possibilities predicted by most current theories of aspect (Beck 2003b). UNT lacks both nominal case and prepositions but allows up to five non-oblique objects, all of which appear to be equally ranked for the purposes of syntactic processes (Beck 2006b, 2007b). The documentation of these and other features will provide theorists with an important source of challenging data from a language whose properties are markedly different from those familiar to linguists working on more commonly-described linguistic families.
The same holds true for the study of language acquisition. 27 hours of naturalistic child language data (three children aged 2;6 – 3;0, taped bi-weekly over a three month period) have been collected and tran-scribed by a PhD student, Vianey Varela, and an additional 10 hours of video were made of children of various ages performing experimental tasks. Current studies of language acquisition are based almost entirely on data from European languages, and many claims about universal patterns and sequences in acquisition and how they are subsumed by the ontological development of human cognition are untested with languages and cultures outside the Western mainstream. On the rare occasion that data from the acquisition of American indigenous languages are brought to bear on these claims, the data usually challenge or refute them (e.g. Allen & Crago 1996; de León 2001). In the case of UNT, for example, the data already show that the most common locative structure used by adults is not a productive part of the language of children before the age of 10 (Varela & Klint 2006; Varela 2007), while most theories assert that the major structures of a language are acquired at a much earlier age. Since the intergenerational transfer of the language in a naturalistic form is now a minority choice, the acquisition of the language will vanish long before the language itself does, which makes documenting acquisition an imperative component of this research programme.
Bibliography
