MORRIS SWADESH: CRITICAL ESSAY

Like his teacher Edward Sapir, Morris Swadesh was a prolific data-gatherer and avid student of languages. From his Russian parents he learned Yiddish. As an undergraduate at the University of Chicago, he studied German and French, and as a graduate and postgraduate at Yale University he concentrated on Nootka, a Canadian indigenous language. On his regular field trips through Canada, the United States, and Mexico, he collected data on more than 20 native languages. In Mexico, he developed programs for indigenous people to attain literacy in their own languages. Later, during World War II, he worked for the war department, editing dictionaries, providing linguistic analyses of foreign languages, and developing teaching materials for Spanish, Russian, Burmese, and Chinese.

Swadesh did his first theoretical work on phonemic analysis, i.e. the analysis of the sound structure or phonology of languages. Sapir, Leonard Bloomfield, Nikolai Trubetzkoy, and others had already advanced the concept of the “phoneme,” an abstract representation of sound types. Swadesh’s contribution was to develop a set of principles to help the phonologist discover phonemes on the basis of the distribution of sounds in a given language. English, for example, uses different "p"-sounds in pit, upper, and spill. Since the pronunciation is in each case clearly dependent on the exact position within the word, Swadesh suggested that these positional sound variants are in “complementary distribution” and should thus be regarded as instances of the same sound type or phoneme.These principles were later applied to word and sentence elements by Zellig Harris. Distributional analysis thus became a general “discovery procedure” for the basic elements of linguistic structure and has remained an integral part of linguistic methodology to this day.

In his extensive investigations of numerous languages, Swadesh gained an increasing appreciation of apparent lexical and structural similarities in different languages, and his interest in comparative historical linguistics grew. Since his study of nearly extinct languages was conducted with limited resources, he felt the need for a standardized procedure for quickly collecting crucial data yielding clues about language relationships.

To determine whether two given languages are related, historical linguists usually employ the “comparative method.” This means that they attempt to reconstruct an ancestor language on the basis of cognates, i.e. arguably related words from different languages (English hound is a cognate of German Hund “dog”). Since cultural development is always accompanied by lexical innovations, Swadesh--and many linguists before him-- felt that presumably more stable “basic vocabulary” would be the best place to start looking for cognates. The basic vocabulary of a language describes body parts and functions, such as skin, blood, drink, natural phenomena like water, sky, bird, smoke, immediate sense experiences and physical dimensions, such as long, red, cold. While the concept of basic vocabulary had been informally used before, Swadesh made it explicit by drafting a list of 100 word meanings that a field worker investigating any language could use for identifying the basic vocabulary of that language. The use of this list, now generally known as the “Swadesh list,” has drawn criticism from its inception. Many linguists believe that it is impossible to enumerate universal meanings, and that the identification of semantically equivalent words in different languages is often highly problematic. Nevertheless, the list has become a widely used tool in comparative linguistics.

Even though the notion of basic vocabulary was already contentious, Swadesh pressed forward and used his list for lexicostatistics, a quantitative method for measuring the similarity of languages. If the basic vocabulary of one language matches that of another to over 90 percent, Swadesh argued, these languages must be closely related. Most linguists believe that the reconstruction of ancestor languages provides more reliable evidence for the relatedness of languages than statistical analyses, so lexicostatistics continues to be viewed with suspicion. Yet, moderate linguists today concede that the Swadesh list and lexicostatistics may be useful for rough initial investigations or for situations where complete data are simply unavailable--which is, in fact, close to what Swadesh had in mind.

Even more controversially, Swadesh claimed that the “decay” of basic vocabulary could be used for “glottochronology,” the dating of ancestor languages analogous to determining the age of fossils on the basis of radioactive decay. Swadesh came to believe that basic vocabulary decays with a rate of 14 percent over 1000 years, so languages would retain on average about 86 percent of their basic vocabulary over this time span. Thus, if the basic vocabularies of two related languages are found to match by 70 percent, they can be assumed to have developed from a single language that existed approximately 12 centuries before.The assumption that basic vocabulary decay is generally uniform has been largely rejected. If one allows that languages, just like societies, may develop at different rates at different times, the assumption of steady vocabulary decay in particular, and the glottochronological method in general, is seriously undermined.

Swadesh’s name has remained symbolic for lexicostatistics and glottochronology, but his central place in the continuing and highly ideological debate about these and related issues seems to rest on misunderstandings and/or polemically slanted readings. He is, for example, accused of introducing lexicostatistics as a shortcut for investigation, attempting to avoid the hard work of reconstruction. However, Swadesh stated repeatedly that a detailed knowledge of the languages under investigation is crucial, and that other data must be considered. Likewise, Swadesh is misleadingly cited as a supporter of the theory that all languages have developed from a single ancestor (the “monogenesis” theory). He certainly conceded that the instinctive vocalizations of early humanoids may be called a species-specific “language,” but he also surmised that as soon as arbitrary signs--i.e. true words and complex linguistic structures--entered the human repertoire, diversification was the instant result.

These ideas were the focus of his major book, The Origin and Diversification of Language, but he died of a heart attack before he was able to complete it. His arguments thus remained somewhat sketchy, which may explain why Swadesh’s even-handed and careful deliberations tend to be overshadowed by the bolder aspects of his thought.

Biography

Born in Holyoke, Massachusetts, 22 January 1909. BA (1930), MA (1931) for dissertation on Nootka aspect, tutored by Edward Sapir, University of Chicago; followed Sapir to Yale University, New Haven, Connecticut, Ph.D. for work on Nootka semantics, 1933, then work at Yale on synchronic phonological theory and on American English grammar, 1933-37. Assistant Professor, University of Wisconsin--Madison, 1937-39. Moved to Mexico City, there Director, Consejo de Lenguas Indígenas, and Director of Linguistics, Departamento de Asuntos Indígenas, 1939; Professor, Instituto Politécnico Nacional de México, Escuela de Antropología, and Departamento de Asuntos Indigenas, 1939-41. Linguist for the War Department in New York City during World War II; Associate Professor, City University of New York, 1948; lost his teaching appointment and had his passport revoked because of “leftist” views and activities; librarian at the Boas Collection, Library of the American Philosophical Society, Philadelphia, PA, 1949-53; independent field work 1953-56. Moved again to Mexico City, there Professor at the Instituto de Historia, Universidad Nacional Autónoma de México, and the Escuela Nacional de Antropología e Historia, 1956-67. Member, Linguistic Society of America (LSA) in 1931, Life Member in 1937; member of two special interest groups of the LSA, 1939; President of the Linguistic Section of the 29th International Congress of Americanists, 1939; editor of Word, 1946-49. Died in Mexico City,  20 July 1967.

Selected Works

The Expression of the Ending-Point Relation in English, French, and German, 1932

“The Phonemic Principle,” Language 10 (1934)

“A Method for Phonetic Accuracy and Speed,” American Anthropologist 39 (1937)

“Linguistics as an Instrument of Prehistory,” Southwestern Journal of Anthropology 15 (1959)

The Origin and Diversification of Language, 1971

Further Reading

Anttila, Raimo, An Introduction to Historical and Comparative Linguistics, New York: Macmillan, 1972; 2nd edition, as Historical and Comparative Linguistics, Amsterdam and Philadelphia: Benjamins, 1989

Harris, Zellig, Methods in Structural Linguistics, Chicago: University of Chicago Press, 1951; as Structural Linguistics, 1960

Hymes, Dell H., editor, Language in Culture and Society, New York: Harper and Row, 1964

Hymes, Dell H., “Morris Swadesh: From the First Yale School to World Prehistory,” in The Origin and Diversification of Language, by Morris Swadesh, Chicago: Aldine Atherton, 1971

Lamb, Sidney M., and E. Douglas Mitchell, editors, Sprung from Some Common Source: Investigations into the Prehistory of Languages, Stanford, California: Stanford University Press, 1991

Newman, Stanley, “Morris Swadesh (1909-1967),” Language 43 (1967)

© 2000 Philipp Strazny