Superfamily (linguistics)

From Wikipedia, the free encyclopedia

In linguistics, a superfamily is a phyletic unit encompassing several language families. The term is generally used for conjectural families, e.g. the Nostratic, Dené-Caucasian, Altaic, and Proto-World groupings. The term is typically invoked to describe theoretical associations between established language families (e.g. between Turkic, Mongolic, and Tungusic languages, and often – but not always – Korean and Japonic, in the case of Altaic), and has come to carry a connotation of hypothesis, and therefore, controversy. As such, it is rarely applied to accepted groupings like the Indo-European language family. Unlike language families established strictly by the comparative method, superfamilies are normally suggested on the grounds of multilateral comparison, which is a much more intuitive method first suggested by Greenberg. It has been shown by Ringe (Transactions of the American Philosophical Society, 1992) that the Greenbergian multilateral comparison is statistically untenable, however this method, facilitating the establishment of linguistic affiliation to a great extent, still appeals to many and is especially popular among amateurs. A stricter approach to superfamily affiliation is suggested by Nichols (Linguistic Diversity in Space and Time, 1992) and is also statistically backed in her book.

In other words, "superfamily" is commonly used as a pejorative term. It does not designate a particular level of language relationship. For example, Afro-Asiatic and Eurasiatic are both very large language families dating back to the Stone Age, but some linguists would call Afro-Asiatic a "language family" because it is generally accepted and Eurasiatic a "language superfamily" because it is not. The term "superfamily" is also occasionally used in a more neutral sense to characterize any higher-order grouping of languages.

The term "macrofamily" has the same meaning but does not carry any pejorative connotation.