¹ Part of this paper was first presented in July 1960 at the Linguistic Institute (University of Michigan), to Paul Friedrich's seminar in Language and Culture; I am obliged to Friedrich for subsequent comments and discussion. Much of the material here presented is amplified in my doctoral dissertation, Bilingualism and Biculturalism in a Huave Community (ms., Yale University, 1961). The paper and my field work in bilingualism have been discussed with Floyd Lounsbury, Laura Nader, Joan Rubin, Glenn Stairs, Morris Swadesh, Milton Warkentin, Jacqueline Wei, and Robert Weitlaner; additional comments on the paper were made by Einar Haugen and Dell Hymes. I gratefully acknowledge their interest. My field research was made possible through award of a National Science Foundation Fellowship and an appointment to the Organization of American States Fellowship Program; in addition, support was extended by the Yale Department of Anthropology. I am thankful for the aegis of the Instituto Nacional de Antropología e Historia and the Instituto de Historia while carrying out my field investigations in Mexico.

² Uriel Weinreich, Languages in contact (Publications of the Linguistic Circle of New York, No. 1; New York, 1953) ; Einar Haugen, Bilingualism in the Americas: A bibliography and research guide (Publication of the American Dialect Society, No. 26; University, Alabama, 1956).

³ Convergence based on transfer is distinct from the biological and older anthropological concept, which described the independent evolution of typologically similar forms without historical connection, or with only remote connection. Convergence in our sense is a concomitant of culture contact as well as language contact. Such merging is excluded from biological systems : speciation produces infinite bifurcation, but mutual infertility of different species prohibits subsequent biological merging. Biological change is almost totally divergent. Linguistic and cultural change, on the other hand, is rarely if ever free from convergence. The analogue of biological speciation in linguistic change is language and dialect formation; but mutual unintelligibility, unlike mutual infertility, is no block to systemic convergence.

⁴ Haugen 11.

⁵ Einar Haugen, The Norwegian language in America: A study in bilingual behavior 1.7 (Philadelphia, 1953).

⁶ Leonard Bloomfield, Language 476 (New York, 1933).

⁷ Bloomfield 462.

⁸ Weinreich 1.

⁹ Bloomfield 464.

¹⁰ Haugen, Bilingualism 40. In this paper ‘interference’, unspecified, will refer to the total phenomenon of linguistic interference. Subsequent sections will occasionally distinguish between integrated and nonintegrated interference, but this usage will be clear from context or specification.

¹¹ Charles Hockett similarly distinguishes between what he calls the individual and mass effects of borrowing; see A course in modern linguistics, Chap. 47 (New York, 1958).

¹² ‘Interlingual identification occurs when speakers equate items in one language with items in another because of their similarities in shape, distribution, or both.‘—Haugen, Bilingualism 67.

¹³ Haugen, Bilingualism 46, uses the term diaphone to describe ‘interlingually identified variants of phonemes’ and the term diamorph for ‘interlingually identified variants of morphemes or groups of morphemes’.

¹⁴ Haugen, Bilingualism 52, re morphemic importation and morphemic substitution: ‘In the former case a new morpheme is introduced into the language, with a phonemic shape which shows diaphonic identification with a morpheme in the source language. This is what is generally known as a loanword. In the latter case, however, no new morpheme is introduced, but the lexeme borrowed is translated by the redistribution of morphemes already found in the language. For this the writer has suggested the term loanshift.‘

¹⁵ Haugen, Bilingualism 15, generalizes a situation of intimate contact for all the native American languages and the colonial European languages which are geographically and/or socially superimposed : ‘Many native speakers are sufficiently segregated to have only rare contact with the dominant languages of their area. But their political dependence makes it necessary everywhere for some of them to learn the dominant languages. The speakers of the latter seldom find it necessary to learn the native languages.‘

¹⁶ Huave is an American Indian language; its 7000 speakers inhabit the five communities referred to. Dialect divergence is pronounced, and the three dialects have been differentially receptive to interference from Spanish; the dialects are nevertheless mutually intelligible, although there are definite weaknesses in communication which are congruent with the dialect boundaries. The genetic pedigree of Huave awaits further comparative investigation. Excellent preliminary descriptive analyses of Huave have been written by Milton and Clara Warkentin and by Glenn and Emily Stairs. The Warkentins are soon to publish some of their descriptive analyses; the Stairs are working on a lexicon and with text materials. My presentation of Huave forms has been made to conform with the Warkentin's phonemicization (Clara E. Warkentin and Evelyn G. Pike, Huave phonemics, ms., n.d.). Some differences appear in the rendering of tones and vowel length. Further alterations involve recognition of two additional phonemes f, ŋ, and deletion of r. f. occurs in Spanish-derived loan words, [η] occurs in Huave tré ŋ ['trerŋ] 'train' and contrasts with [n] in that VC+ environment, e.g. ndén [_ln'den] ‘shade’; elsewhere, [n] and [ŋ] are in complementary distribution. In my data flap and trill are in free variation.

¹⁷ Zapotec is also an American Indian language. The variety spoken here is the so-called Isthmus ‘dialect’, actually one of a number of mutually unintelligible languages. Although this too is a lower language, the number of speakers is still very great; its displacement by Spanish is slow; and there is some interference in the Isthmus dialect of Spanish from Zapotec. All these facts indicate that Zapotee, although it is a lower language in its relationship to Spanish, enjoys greater prestige than many Indian languages in Mexico, a privilege it shares with Tarascan and Yucatec Maya. Although there is contact between Zapotee and Huave, this is negligible in San Mateo: San Mateños who are bilingual in Zapotee constitute less than 8% of the population, and the amount of Huave-Zapotec interference is minimal. I have noted only eleven instances of Huave-Zapotec-Spanish trilingualism in San Mateo.

¹⁸ See Haugen's discussion of The Age of Learning, Bilingualism 72, 73, and Weinreich's Order of Learning and Age, op. cit. 76–7 and 94–5.

¹⁹ Kenneth Pike and Charles Fries, Coexistent phonemic systems, Lg. 25.29–50 (1949), argue that in Mazatec, Spanish interference sets up a definable environment which may be excluded for the purposes of description.

²⁰ This is the 100-word list which appears in revised form in Morris Swadesh, Towards greater accuracy in lexicostatistic dating, IJAL 21.121–37 (1955).

²¹ Einar Haugen, Language contact, Proceedings of the VIII International Congress of Linguists 783 (Oslo, 1958).

²² Hockett 422.

²³ Diebold, Impresiones particulares sobre el bilingüismo en México (ms., New Haven, 1958).