History of classification Uto-Aztecan has been accepted by linguists as a language family since the early 1900s, and six subgroups are generally accepted as valid:
Numic,
Takic, Pimic,
Taracahitic,
Corachol, and
Aztecan. That leaves two ungrouped languages:
Tübatulabal and
Hopi (sometimes termed "
isolates within the family"). Some recent studies have begun to question the unity of Taracahitic and Takic and computer-assisted statistical studies have begun to question some of the long-held assumptions and consensuses. As to higher-level groupings, disagreement has persisted since the 19th century. Presently scholars also disagree as to where to draw language boundaries within the
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.
Daniel Garrison 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. 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 Uto-Aztecan corresponds to Powell's "Shoshonean", and the latter is all the rest: Powell's "Sonoran" plus Aztecan. Northern Uto-Aztecan was proposed as a genetic grouping by
Jeffrey Heath in based on morphological evidence, and
Alexis Manaster Ramer in adduced phonological evidence in the form of a sound law.
Terrence Kaufman in 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.
Wick R. 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 and . Reviewing the debate, 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. also considered the north–south split to be valid based on phonological evidence, confirming both groupings. adduced further evidence for the unity of Southern Uto-Aztecan as a valid grouping. 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. ).
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 . 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.
Proposed external relations An "Aztec–Tanoan" macrofamily that unites the Uto-Aztecan languages with the
Tanoan languages of the southwestern United States was first proposed by
Edward Sapir in the early 20th century, and later supported with potential lexical evidence by other scholars. This proposal has received much criticism about the validity of the proposed cognate sets and has been largely abandoned since the end of the last century as unproven. ==Proto-Uto-Aztecan==