Uto Aztecan languages

Proto-language and Uto-Aztecan homeland

The Proto-Uto-Aztecan language is the hypothetical common ancestor of the Uto-Aztecan languages. Authorities on the history of the language group have usually placed the Proto-Uto-Aztecan homeland in the border region between the United States and Mexico, namely the upland regions of Arizona and New Mexico and the adjacent areas of the Mexican states of Sonora and Chihuahua, roughly corresponding to the Sonoran Desert and the western part of the Chihuahuan Desert. The proto-language would have been spoken by Mesolithic foragers in Aridoamerica, about 5,000 years ago.

Based on clues to the ecological niche inhabited by the proto-Uto-Aztecans offered reconstructions of the plant related vocabulary, Fowler placed the center of the proto-Uto-Aztecan dialect continuum in Central Arizona with Northern dialects extending into Nevada and the Mojave desert, and Southern dialects extending south through the Tepiman corridor into Mexico. The homeland of the Numic languages has been placed in Southern California near Death Valley, and the homeland of the proposed Southern Uto-Aztecan group has been placed on the coast of Sonora.

A contrary proposal, that suggests the homeland of Proto-Uto-Aztecan to have been much further to the south, was published in 2001 by Jane H. Hill, based on her reconstruction of maize-related vocabulary in Proto-Uto-Aztecan. By her theory, the assumed speakers of Proto-Uto-Aztecan were maize cultivators in Mesoamerica, who gradually moved north, bringing maize cultivation with them, during the period of roughly 4,500 to 3,000 years ago. The geographic diffusion of speakers corresponded to the breakup of linguistic unity. This hypothesis has been criticized on several grounds, and it is not generally accepted by Uto-Aztecanists. A survey of agriculture-related vocabulary by Merrill (2012) found that the agricultural vocabulary can only be reconstructed for Southern Uto-Aztecan. This supports a conclusion that the Proto-Uto-Aztecan speech community did not practice agriculture, but only adopted it after entering Mesoamerica from the North.

A recent proposal by David L. Shaul presents evidence suggesting contact between proto-Uto-Aztecan and languages of central California such as Esselen and the Yokutsan languages. This leads Shaul to suggest that proto-Uto-Aztecan was spoken in California's Central Valley area, and formed part of an ancient Californian linguistic area.

Geographic distribution

Uto-Aztecan languages are spoken in the North American mountain ranges and adjacent lowlands of the western United States (in the states of Oregon, Idaho, Montana, Utah, California, Nevada, Arizona) and of Mexico (states of Sonora, Chihuahua, Nayarit, Durango, Zacatecas, Jalisco, Michoacán, Guerrero, San Luis Potosí, Hidalgo, Puebla, Veracruz, Morelos, Estado de México, and the Federal District). Classical Nahuatl, the language of the Aztecs, and its modern relatives are part of the Uto-Aztecan family. The Pipil language, an offshoot of Nahuatl, spread to Central America by a wave of migration from Mexico, and formerly had many speakers there. Now it has gone extinct in Guatemala and Honduras, and it is nearly extinct in western El Salvador, all areas dominated by use of Spanish.

History of classification

Uto-Aztecan has been accepted by linguists as a language family since the early 1900s, and six subgroups are accepted as valid by all experts: Numic, Takic, Pimic, Taracahitic, Corachol, and Aztecan. This leaves two ungrouped languages—Tübatulabal and Hopi (sometimes termed "isolates within the family"). As to higher-level groupings, disagreement has persisted since the 19th century. Presently scholars also disagree as to where to draw language boundaries within dialect continua.

The similarities among the Uto-Aztecan languages were noted as early as 1859 by J.C.E. Buschmann, but he failed to recognize the genetic affiliation between the Aztecan branch and the rest. He ascribed the similarities between the two groups to diffusion. Brinton added the Aztecan languages to the family in 1891 and coined the term Uto-Aztecan. John Wesley Powell, however, rejected the claim in his own classification of North American indigenous languages (also published in 1891). Powell recognized two language families: "Shoshonean" (encompassing Takic, Numic, Hopi, and Tübatulabal) and "Sonoran" (encompassing Pimic, Taracahitan, and Corachol). In the early 1900s Alfred L. Kroeber filled in the picture of the Shoshonean group, while Edward Sapir proved the unity among Aztecan, "Sonoran", and "Shoshonean". Sapir's applications of the comparative method to unwritten Native American languages are regarded as groundbreaking. Voegelin, Voegelin & Hale (1962) argued for a three way division of Shoshonean, Sonoran and Aztecan, following Powell.

As of about 2011, there is still debate about whether to accept the proposed basic split between "Northern Uto-Aztecan" and "Southern Uto-Aztecan" languages. Northern-Utoaztecan corresponds to Powell's "Shoshonean", while the latter is all the rest, i.e., Powell's "Sonoran" plus Aztecan. Northern Uto-Aztecan was proposed as a genetic grouping by Jeffrey Heath (1978) based on morphological evidence, and Manaster Ramer (1992) adduced phonological evidence in the form of a sound law. Kaufman (1981) accepted the basic division into Northern and Southern branches as valid. Other scholars have rejected the genealogical unity of either both nodes or the Northern node alone. Miller's argument was statistical, arguing that Northern Uto-Aztecan languages displayed too few cognates to be considered a unit. On the other hands he found the number of cognates among Southern Uto-Aztecan languages to suggest a genetic relation. This position was supported by subsequent lexicostatistic analyses by Cortina-Borja & Valiñas-Coalla (1989) and Cortina-Borja, Stuart-Smith & Valiñas-Coalla (2002). Reviewing the debate, Haugen (2008) considers the evidence in favor of the genetic unity of Northern Uto-Aztecan to be convincing, but remains agnostic on the validity of Southern Uto-Aztecan as a genetic grouping. Hill (2011) also considered the North/South split to be valid based on phonological evidence, confirming both groupings. Merrill (2013) adduced further evidence for the unity of Southern Uto-Aztecan as a valid grouping.

Hill (2011) also rejected the validity of the Takic grouping decomposing it into a Californian areal grouping together with Tubatulabal.

Some classifications have posited a genetic relation between Corachol and Nahuan (e.g. Merrill (2013)). Kaufman recognizes similarities between Corachol and Aztecan, but explains them by diffusion instead of genetic evolution. Most scholars view the breakup of Proto-Uto-Aztecan as a case of the gradual disintegration of a dialect continuum.

Present scheme

Below is a representation of the internal classification of the language family based on Shaul (2014). The classification reflects the decision to split up the previous Taracahitic and Takic groups, that are no-longer considered to be valid genetic units. Whether the division between Northern and Southern languages is best understood as geographical or phylogenetic is under discussion. The table contains demographic information about number of speakers and their locations based on data from The Ethnologue. The table also contains links to a selected bibliography of grammars, dictionaries on many of the individual languages.(^† = extinct)

In addition to the above languages for which linguistic evidence exists, it is suspected that among dozens of now extinct, undocumented or poorly known languages of northern Mexico, many were Uto-Aztecan.

Extinct languages

A large number of languages known only from brief mentions are thought to have been Uto-Aztecan languages, that became extinct before being documented.

Vowels

Proto-Uto-Aztecan is reconstructed as having an unusual vowel inventory: *i *a *u *o *ɨ. Langacker (1970) demonstrated that the fifth vowel should be reconstructed as *ɨ as opposed to *e—there had been a long-running dispute over the proper reconstruction.

Consonants

*n and *ŋ may have actually been *l and *n, respectively.