Professor Joseph Greenberg was one of the most important
linguists ever. He revolutionized the
classification of languages in more ways than one and founded one of the main fields of
linguistics, that of "
typology and universals".
The formal methods available for establishing language families had been used since about 1800, and many families had been firmly established. This can't be done indefinitely far back because with sufficient depth of time, surviving common inheritances from an ancestral language in two distant branches could be equally well due to chance. Greenberg however did establish one major phylum or superfamily and believed in a number more, which linguists have not (yet?) accepted.
In 1955 he published a classification of African languages, expanding it in his 1963 book The Languages of Africa. He claimed that all native languages (i.e. excluding Malagasy and Afrikaans, both of which were obvious newcomers from elsewhere) belonged to one of just four families: Niger-Kordofanian, Afro-Asiatic, Nilo-Saharan, and Khoisan.
Niger-Kordofanian covers the great majority of them, all over west, central, and southern Africa, including the large Bantu family (Swahili, Zulu, etc.). Afroasiatic is mostly north of the Sahara, and also includes the Semitic languages of Asia. Nilo-Saharan is a fairly small scattered group in the east between those two major groups; and Khoisan is all those of the so-called Bushmen and Hottenots in the south. This scheme of Greenberg's was strongly resisted at first but is now universally admitted.
His most controversial claim is not accepted by a majority of linguists but could still turn out to be true: that almost all the languages of the Americas belong to a single superfamily, which he called Amerind, and which reflects the settlement of the continents from across the Bering Strait. Excluded from this group are two other families in the north, Eskimo-Aleut and Na-Dene (the latter includes Navaho). These he said represented later crossings of Beringia. Untangling the truth of all this is one of the most interesting topics in linguistics, archaeology, anthropology, and human evolution today. His 1986 paper co-written with two anthropologists Christy Turner and Stephen Zegura was called "The Settlement of the Americas: a comparison of linguistic, dental, and genetic evidence". He wrote a book Language in the Americas in 1987.
He also proposed an Indo-Pacific phylum, covering the New Guinean (Papuan) languages together with Andamanese and Tasmanian.
Greenberg's last major proposal was Eurasiatic, similar to the existing Nostratic theory but with different emphases. He grouped Indo-European (English, Latin, Russian, Hindi etc.) with many of the smaller groups of northern Asia, crossing over into North America to include Eskimo-Aleut.
Greenberg relied for the more widespread groupings on a technique called mass comparison, in which large groups of words are compared to pick out resemblances that have survived the history of the group. This is not generally accepted, because it violates the more conservative principle of the comparative method, which tries to explain each difference by a rule. Many of the arguments against mass comparison are statistical, saying that such resemblances could be expected by chance anyway.
His 1960 paper "Some Universals of Grammar with Particular Reference to the Order of Meaningful Elements" is one of the most important in linguistics.* He discovered that adjective versus noun order, preposition versus postposition, order of verb and subject and object, and other such features, often possessed strong correlations among each other. See my typology node for greater detail. This discovery replaced the old view of languages as being synthetic, analytic, or isolating.
Joseph Harold Greenberg was born in New York on 28 May 1915. He also had an aptitude for music and considered a career as a concert pianist. He got a PhD at Northwestern Univeristy in 1940, studied and taught at Columbia University and took every language course he could, and did fieldwork among the Hausa of Nigeria. He moved to Stanford in 1962, retired in 1986, and continued to study and write almost to the end of his life. He died in Stanford on 7 May 2001.
* Reprinted 1966 in Universals of Language, MIT Press, pp. 73-113. Constantly quoted ever since.