The primary language of the United States is English, enriched by words borrowed from the languages of Indians and immigrants, predominantly European. Spanish is also spoken by a sizable minority.
When European settlement began, Indians living north of Mexico spoke about 300 different languages now held to belong to 58 different language families. Only 2 such families have contributed noticeably to the American vocabulary: Algonkian in the Northeast and Aztec-Tanoan in the Southwest. From Algonkian languages, directly or sometimes through Canadian French, English has taken such words as moose, skunk, caribou, opossum, woodchuck, and raccoon for New World animals; hickory, squash, and tamarack for New World flora; and succotash, hominy, mackinaw, moccasin, tomahawk, toboggan, and totem for various cultural items. From Nahuatl, the language of the Aztecs, terms such as tomato, mesquite, coyote, chili, tamale, chocolate, and ocelot have entered English, largely by way of Spanish. A bare handful of words come from other Indian language groups, such as tepee from Dakota Siouan, catalpa from Creek, sequoia from Cherokee, hogan from Navaho, and sockeye from Salish, as well as cayuse from Chinook.
Professional dialect research, initiated in Germany in 1878 and in France in 1902, did not begin in the United States until 1931, in connection with the Linguistic Atlas of New England (1939–43). This kind of research, requiring trained field-workers to interview representative informants in their homes, subsequently was extended to the entire Atlantic Coast, the north-central states, the upper Midwest, the Pacific Coast, the Gulf states, and Oklahoma. The New England atlas, the Linguistic Atlas of the Upper Midwest (1973–76), and the first two fascicles of the Linguistic Atlas of the Middle and South Atlantic States (1980) have been published, along with three volumes based on Atlantic Coast field materials. Also published or nearing publication are atlases of the north-central states, the Gulf states, and Oklahoma. In other areas, individual dialect researchers have produced more specialized studies. The definitive work on dialect speech, the American Dialect Society's monumental Dictionary of American Regional English, began publication in 1985.
Dialect studies confirm that standard English is not uniform throughout the country. Major regional variations reflect patterns of colonial settlement, dialect features from England having dominated particular areas along the Atlantic Coast and then spread westward along the three main migration routes through the Appalachian system. Dialectologists recognize three main dialects—Northern, Midland, and Southern—each with subdivisions related to the effect of mountain ranges and rivers and railroads on population movement.
The Northern dialect is that of New England and its derivative settlements in New York; the northern parts of Ohio, Indiana, Illinois, and Iowa; and Michigan, Wisconsin, northeastern South Dakota, and North Dakota. A major subdivision is that of New England east of the Connecticut River, an area noted typically by the loss of /r/ after a vowel, and by the pronunciation of can't, dance, half, and bath with a vowel more like that in father than that in fat. Generally, however, Northern speech has a strong /r/ after a vowel, the same vowel in can't and cat, a conspicuous contrast between cot and caught, the /s/ sound in greasy, creek rhyming with pick, and with ending with the same consonant sound as at the end of breath.
Midland speech extends in a wide band across the United States: there are two main subdivisions, North Midland and South Midland. North Midland speech extends westward from New Jersey, Delaware, and Pennsylvania into Ohio, Illinois, southern Iowa, and northern Missouri. Its speakers generally end with with the consonant sound that begins the word thin, pronounce cot and caught alike, and say cow and down as /caow/ and /daown/. South Midland speech was carried by the Scotch–Irish from Pennsylvania down the Shenandoah Valley into the southern Appalachians, where it acquired many Southern speech features before it spread westward into Kentucky, Tennessee, southern Missouri, Arkansas, and northeast Texas. Its speakers are likely to say plum peach rather than clingstone peach and snake doctor rather than dragonfly.
Southern speech typically, though not always, lacks the consonant /r/ after a vowel, lengthens the first part of the diphthong in write so that to Northern ears it sounds almost like rat, and diphthongizes the vowels in bed and hit so that they sound like /beuhd/ and /hiuht/. Horse and hoarse do not sound alike, and creek rhymes with meek. Corn bread is corn pone, and you-all is standard for the plural.
In the western part of the United States, migration routes so crossed and intermingled that no neat dialect boundaries can be drawn, although there are a few rather clear population pockets.
The majority of Spanish speakers live in the Southwest, Florida, and eastern urban centers. Refugee immigration since the 1950s has greatly increased the number of foreign-language speakers from Latin America and Asia.
Very early English borrowed from neighboring French speakers such words as shivaree, butte, levee, and prairie; from German, sauerkraut, smearcase, and cranberry; from Dutch, stoop, spook, and cookie; and from Spanish, tornado, corral, ranch, and canyon. From various West African languages, blacks have given English jazz, voodoo, and okra.
Educational problems raised by the presence of large blocs of non-English speakers led to the passage in 1976 of the Bilingual Educational Act, enabling children to study basic courses in their first language while they learn English. A related school problem is that of black English, a Southern dialect variant that is the vernacular of many black students now in northern schools.