3.

The Sociocultural Background of Dialect Areas in American English

3.1 Introduction

A brief description of several attempts to achieve insight into the dialectal structure of an area on the basis of known heteroglossic lines, or to provide at least a tentative orientation, should throw light on some of the problems that must be faced.

In order to characterize and to evaluate the proposed schemes, the following factors must be taken into consideration:

(1) The character and fullness of the available data as determined by the content of the questionnaire, the choice of communities, and the method of gathering the data.

(2) The method of establishing boundaries within the area: (a) on the basis of congruent bundles of lexical, phonological, and morphological heteroglosses; (b) on the basis of bundles of any of the three types taken separately; or (c) on the basis of representative or diagnostic heteroglosses. (d) A boundary can also be suggested “intuitively” by a “convenient” heterogloss.

(3) Whether a diagnostic heterogloss is structural or nonstructural is another point that must be taken into account.

3.2 The Dialectal Structure of New England

Phonological and lexical heteroglosses form loose bundles running northward from Long Island Sound to Canada. These bundles divide New England into an Eastern and a Western dialect area (see Figure 9).

The course of these strands of heteroglosses exhibits no congruence with the boundaries of the New England states. Connecticut, Massachusetts, and Vermont are cut in half, while New Hampshire and Maine lie well within the Eastern area. Political boundaries have evidently not seriously interfered with communication between the inhabitants of the several states.

When we turn to a map showing the distribution of the population in New England (1930), we discover that of the two areas of concentration one lies to the east of the bundles of heteroglosses and the other to the west (see Figure 10). From this demographic fact we anticipate that the development of these two major dialect areas of New England is somehow connected with these two population centers and their history.

FIGURE 9: New England: Eastern Pronunciations. From Kurath, Handbook of the Linguistic Geography of New England.

We also observe that most of the heteroglosses run through sparsely settled areas—along the Green Mountains of Vermont and through the hill country of central Massachusetts and eastern Connecticut. Such areas impede communication and thus tend to create and to stabilize existing dialect boundaries.

The Eastern concentration area extends from Narragansett Bay (Providence, R. I.) to Casco Bay (Portland, Me.). Its dominant center is Massachusetts Bay (Boston). The Western area of concentration extends from Long Island Sound (New Haven) to Hartford and Springfield on the Connecticut River. Both of these centers date from early Colonial times and have continuously, and increasingly, influenced the speech of their back county.

FIGURE 10: New England: Population Distribution, 1930

From J. K. Wright, editor, New England’s Prospect: 1933, page 23. New York, 1933.

The chronology of the settlement of New England has fortunately been worked out in considerable detail. Between 1630 and 1645 colonies were planted along the open Atlantic and on the lower Connecticut River. Expanding slowly, they did not establish contact with each other for about half a century (see Figure 11), a period during which some regional differences in usage must have become established. Half a century later (1725) these two settlement areas were in contact with each other in Connecticut and in parts of Massachusetts, but the sparsely settled hill country connecting them interfered with communication.

FIGURE 11: New England: Chronology of Settlement

Cities: B(oston, N(ew Haven, P(rovidence, S(pringfield

Based upon Kurath, Handbook of the Linguistic Geography of New England (foldout map).

Even more important is the fact that each of these areas developed a transportation system radiating from the old centers outward to the peripheral settlements, which served to consolidate them economically, socially, and culturally. This process of integration continued with increasing force in the nineteenth century when turnpikes and railroads were built. The network of railroads in operation in 1930 effectively portrays this situation (see Figure 12). Add to this the prestige enjoyed by Boston-Cambridge-Concord from the middle of the nineteenth century onward in the fields of literature and scholarship, and you have a picture of the sociocultural forces that have preserved Eastern New England as a highly distinctive dialect area.

FIGURE 12: New England: Railroads

From J. K. Wright, editor, New England’s Prospect: 1933, page 345. New York, 1933.

Eastern New England is of course not a wholly uniform speech area. Settlement history is reflected, for instance, in the survival of tempest “storm” and cade “pet lamb” from Narragansett Bay to Cape Cod, i.e., in the colonies established in Rhode Island and in the Plymouth area [Kurath 1949: Figure 12]. Recession of the highly distinctive checked vowel / e /, as in coat, road, home, is in progress, as shown by its regional, social, and age distribution [Avis 1961].

This brief account of the sociocultural background of the two major speech areas of New England must suffice. Its present purpose is to illustrate the method by which the discovery of congruences between dialect boundaries and the areas set off by them and sociocultural boundaries and domains leads to more or less plausible historical interpretations of areal linguistic phenomena.

3.3 The Boundary between the Northern and the Midland Dialect Areas

A close-knit bundle of lexical, phonological, and morphological heteroglosses runs through northern Pennsylvania, some thirty miles south of the state line, and continues westward into Ohio. At the eastern end (near Scranton) the heteroglosses fan out, some of them swerving southeastward through New Jersey, others continuing eastward to pass north of New York City (see Figure 13).

The prominent dialect boundary in northern Pennsylvania clearly reflects settlement history, channeled and reinforced by the topography. While the Pennsylvania settlements expanded up the Susquehanna Valley to the forested area in the northern part of the state, westward across the Alleghenies to the upper Ohio and its tributaries (Pittsburgh), and southwestward into the valleys of the Appalachians, New Englanders migrated westward into the basin of the Great Lakes, skirting the Dutch settlements in the Hudson Valley. New Englanders also moved southward into eastern New Jersey. From Upstate New York, some of them pushed southward into the wooded hill country of northern Pennsylvania. These lines of expansion are suggested in Figure 14. The chronology of the settlement is shown in Paullin- Wright 1933: Plate 76A-E.

In the Hudson Valley and in New Jersey the transition area between the North and the Midland reflects partly the complicated history of the settlement—English over Dutch—and partly the later effects of the chief communication route (New York-Philadelphia) that intersects the settlement boundary between East Jersey and West Jersey.

The speech of Upstate New York is obviously derived from that of New England. Features peculiar to eastern New England, such as the loss of postvocalic / r / in beard, hard, board, or the low front vowel /a / in half, pass, aunt, survive only sporadically. The chief reason for this is the fact that the majority of the settlers came from rural western New England at a time when the increasing population of eastern New England was absorbed into the developing industries and engaged in seafaring.

In Pennsylvania the dialect boundary between the North and the Midland, created by the population movements outlined above, was supported in later years by the sparsely settled belt in the heavily forested section of northern Pennsylvania.

FIGURE 13: The Southern Boundary of Three Northern Words

From H. Kurath, A Word Geography of the Eastern States.

The westward extension of this boundary has been established on the basis of the sampling survey of the Great Lakes area directed by A. H. Marckwardt (see Figure 15). On lexical evidence, the dividing line between the North and the Midland runs through northern Ohio, Indiana, and Illinois to the Mississippi River (near Burlington, Iowa). To the north of it, New England regionalisms are current, of which many can be traced all the way to the North Pacific Coast, as C. E. Reed [1956, 1957] has shown.

3.4 The Structure of the Upper South

The Upper South, focused on eastern Virginia, constitutes a rather well defined subdivision of the Southern dialect area. Its northern boundary runs in an arc from the Atlantic Ocean through Delaware (north of Dover) and Maryland (north of Baltimore) to the Blue Ridge Mountains of Virginia. From there the dividing line follows the Blue Ridge southwestward to the upper reaches of the Roanoke River. To the north of this boundary lies the North Midland area, to the west of it the South Midland area. The southern limit of this area is less clearly defined. It has the character of a transition belt formed by spaced heteroglosses, some of which dip into north central North Carolina, while others follow the tidal inlet of the James River.

FIGURE 14: The Boundary between the Northern and the Midland Dialect Areas

The arrows suggest the westward expansion of the settlements from New England and from eastern Pennsylvania in the latter part of the eighteenth and the first two decades of the nineteenth century.

The focal area of the Upper South is suggested by three concentrie lexical isoglosses shown in Figure 16 and by an important phonological feature in Figure 17.

The search for extralinguistic factors that might be responsible for the prominent dialect boundary that sets the Upper South off from the Midland reveals immediately that there is no congruence whatever with major political boundaries, i.e., with state lines: the linguistic boundary cuts right through Delaware, Maryland, and Virginia. Secondly, although this boundary rests upon the Blue Ridge Mountains in Virginia, which rise steeply out of the coastal plain to form a rather formidable natural barrier to communication, it runs right through the coastal plain in Maryland and Delaware. An investigation of settlement history, the character of the economy, and the social organization of the Upper South readily leads to the discovery of forces that have shaped this dialect area.

FIGURE 15: The Great Lakes Area: Northern Words

From A. H. Marckwardt, Principal and Subsidiary Dialect Areas in the North-Central States.

FIGURE 16: The Virginia Piedmont: Word Boundaries

From H. Kurath, A Word Geography of the Eastern States.

The Upper South was settled by gradual expansion from the colonies planted on Chesapeake Bay in the seventeenth century. These movements up the river valleys were controlled to a large extent by the plantation aristocracy engaged in the cultivation of tobacco for the European market. As the fertility of the “old fields” was exhausted, new land suitable for growing tobacco was cleared. The heart of this plantation country is strikingly reflected in the concentration of Negro slaves in the Piedmont of Virginia from 1790 to 1860 [Paullin-Wright 1932: plates 67B and 68B].

From the colonial seaports of this area—Richmond on the James River, Fredericksburg on the Rappahannock, and Alexandria on the Potomac—the tobacco was shipped to England, and in return manufactured goods were imported. The common interest of the dominant plantation aristocracy consolidated the area both economically and socioculturally. The linguistic integration of this area clearly emerged from this situation.

The areas north and west of the Upper South were settled largely from, and by way of, Pennsylvania. Northern Maryland and the Valley of Virginia (west of the Blue Ridge) never were plantation country. The settlers, many of them Ulster Scots and Germans, engaged in general farming; wheat fields and orchards characterized the landscape. In Virginia, the conflicting economic interests between the coastal plain and the Valley of Virginia, and consequently the divergent attitudes toward slavery, created a regional antagonism that tended to keep the two dialect areas apart.

FIGURE 17: The Virginia Piedmont: Postvocalic / /

From H. Kurath and R. I. McDavid, Jr., The Pronunciation of English in the Atlantic States.

In Figure 18 the settlement paths of the South and of the South Midland, and the dialect boundary separating them, are presented schematically.

Although set off from the Midland dialect area by a well defined boundary, especially in its northern sector, the Upper South is still rather far from uniform in its linguistic usage. The Eastern Shore of Chesapeake Bay has preserved a considerable number of old local features. Even the points of land between the tidal inlets of the rivers on the western shore have not yet been brought fully into line with the dominant focal area of the Upper South—the Piedmont of Virginia. Diffusion from this center is still in progress, as evidenced by the social dissemination of variants on its periphery [Kurath 1964: 135–44].

FIGURE 18: Settlement Paths of the Coastal South and of the South Midland

3.5 The Lower South

The sociocultural factors underlying the dialect situation of the Lower South, of which South Carolina is the focus, resemble those of the Upper South in many ways. There are, however, some rather marked differences.

The plantation economy (indigo, rice, and later cotton) controlled the life of the people from the very beginning and created the steeply graded social structure of the Lowcountry. Even more than in Virginia, the planter class dominated economic, political, and social affairs and in time largely submerged the Upcountry, where Ulster Scots and Germans from Pennsylvania had settled in considerable numbers. This development has been admirably described by a South Carolinian, R. I. McDavid, Jr. [1948]. The author shows how some linguistic features once peculiar to the Lowcountry of South Carolina have been diffused to the Upcountry from the prestige diale et of the planter class. To be sure, some other distinctive traits of Lowcountry speech have not spread inland, as the ingliding articulation of the vowels /e, o / of eight [ e^ə t], coat [ ko^ə t] in the Charleston dialect, a feature that seems to be receding even there.

Some phonological features peculiar to the Lower South are displayed in Kurath-McDavid 1961: maps 16, 19, 21, 25, 156.

3.5 The Pacific States

The English spoken in the Pacific States־־־California, Oregon, and Washington—has been investigated by David W. Reed and by Carroll E. Reed, partly by direct observation in the field and partly by correspondence. Their questionnaire includes a fair number of lexical items for which regional synonyms had been established in the Eastern States [Kurath 1949] and in the derivative North Central States [Davis 1948].

The evidence secured by correspondence in response to a “check list” of words whose currency is regionally restricted in the Eastern States is presented and analysed by D. Reed [1954] for California and by C. Reed [1956, 1957] for Oregon, Washington, and Idaho.

The chief findings are: (1) In this vast area, extending some eleven hundred miles from Mexico to Canada, the dissemination of regional Eastern words differs little from state to state, although San Francisco Bay, the Willamette Valley of Oregon, and the Puget Sound area of Washington were separate “growing points.” (2) Regional synonyms of the East are current side by side without any marked differences in their relative frequency from state to state. (3) Words used in large sections of the Atlantic Slope and/or the North Central States appear with greater frequency than synonyms restricted to subareas of the East. (4) Regional words derived from the Northern dialect area (New England and the basin of the Great Lakes) and from the North Midland (Pennsylvania and the Ohio Valley) appear with similar frequency, whereas words peculiar to the South Atlantic States are rare. (5) The distribution by age groups shows that regional words are receding sharply all the way from California to Washington.

The authors point out some of the sociocultural factors underlying the behavior of the words investigated. Other factors can easily be adduced, since the settlement history of the Pacific Coast and later developments are so well known.

The salient facts can be briefly stated. Beginning with the 1840’s, California, Oregon, and Washington received their rapidly growing English-speaking population largely from the same sources: New England, the Middle Atlantic States, and the Midwest. Immigrants from abroad poured into all of the Pacific States, though California had the largest share. In time the foreigners learned the language of the dominant English stock. These sociocultural factors account for the striking similarity in word usage throughout the Pacific States.

As in other sections of the United States, the recession of the largely rural regional words reflects the mechanization of farming and the rapid urbanization since the turn of the century. Nearly three-fourths of the population of the Pacific States now live in urban areas and are quite unfamiliar with life on the farm, where regional words brought in from the East survive as valuable evidence for the provenience of the native American stock of the Far West.

Among the words that clearly establish the Northern provenience of large elements among the English-speaking settlers in the Pacific States are the following: co boss! a call to cows; angleworm “earthworm”; darning needle “dragon fly.” Confined to the Northern dialect area on the Atlantic Slope [Kurath 1949: Figures 99, 140, 141] and to the basin of the Great Lakes [Davis 1948: Maps 163, 165, 239], these lexical items unmistakably reflect the westward migration of New Englanders and their descendants in the basin of the Great Lakes to the Pacific Coast.

The westward trek from Pennsylvania and the Ohio Valley—the North Midland speech area—is shown with equal clarity by the currency of regional Midland words in the Pacific States, among them blinds “roller shade,” greenbeans “string beans,” sook! a call to cows, for which see Kurath 1949: Figures 49, 133, 99 and Davis 1948: Maps 178, 210, 163.

The relative frequency of these Northern and Midland words appears to be much the same in California and in Washington, which suggests that the proportion of “Northerners” and “Midlanders” in these two states as a whole differs little.

The behavior of regional words common to the North and the North Midland of the Eastern States supports this inference. In addition, it confirms the rule that the frequency with which such words appear in the “derivative area” of the Pacific States depends upon the extent of their currency in the “mother area.” Thus whinny [Kurath 1949: Figure 97; Davis 1948: Map 161] and skunk [Kurath 1949: Figure 137], shared by the North and the North Midland, are much more widely disseminated in California and Washington than the Northern and the Midland words illustrated above. In conformity with this rule, words confined to subareas of the North or the Midland survive only in scattered instances or not at all. See the lists in D. Reed 1954: 13-14 and C. Reed 1956: 7.

The calculation of the frequency with which regional words imported from the East occur in the Pacific States is an important contribution of the Reeds to the method of area linguistics, which has been applied with significant results to the regional vocabulary of Texas by E. B. Atwood [1962]. This device has special importance for dealing with derivative speech areas in which usage reflects recent mixture without any clear regional dissemination of the variants.

3.7 Texas

E. B. Atwood’s book on The Regional Vocabulary of Texas [1962] is a well planned and circumspect treatment of one aspect of the linguistic situation in the state of Texas and its neighbors—Louisiana, Arkansas, Oklahoma, and New Mexico. Most of his data were gathered between 1950 and 1960 by his friends and students with the help of a questionnaire. For Oklahoma, W. R. Van Riper contributed his field materials, for Louisiana, M. Babington.

For his questionnaire Atwood selected (1) items referring to farming and to life in the countryside for which regionally restricted (varying) terms are current in the eastern United States [Kurath 1949]; (2) designations referring to the range cattle industry and the topography of the western part of the state; and (3) some expressions that seemed to be peculiar to Texas or to southern Louisiana. The last two groups enabled him to show the Spanish and the French contributions to the vocabulary of Texas and to set off the areas in whicl they are current. The stock of regional terms brought in from the “mother” area of American English along the Atlantic during the westward movement, and recent trends in their currency, naturally are his primary concern.

The author found that most of the English regional terms used (or formerly used) by Texans are derived from the South Atlantic States—the Southern and the South Midland dialect areas [Kurath 1949: Figure 3], and that the extent to which they are current in Texas stands in a fairly clear relation to the size of the areas they occupied in the Eastern States around 1940.

Thus bucket “pail,” Christmas gift “merry Christmas,” (corn) shucks “husks,” dog irons ~ fire dogs “andirons,” paling “picket” (of a fence), pullybone “wishbone,” which are common to the South and the South Midland on the Atlantic slope, are widely used in Texas; and so are nicker “whinny” and clabber cheese “cottage cheese,” although the former is not current in the Carolinas and the latter is restricted to the South Midland and North Carolina.

Some terms that in the East are confined to the Southern speech area (the plantation country of the coastal plain), as snap beans “greenbeans,” carry home “take (somebody) home,” and low “moo” have (or had until recently) considerable currency in Texas and adjoining areas; so do sook! a call to cows, (quarter) till (eleven) “to,” and the phrase (I) want off “want to get off,” which are peculiar to Pennsylvania and the South Midland (West Virginia to the Upcountry of the Carolinas). See Figure 19.

Words that have limited currency in the East are infrequent in Texas or not used at all: Virginian batter-bread “soft corn cake”; coastal Carolinian spider “frying pan,” press peach “clingstone peach,” whicker “whinny”; and fire-board “mantle shelf,” red-worm “earth worm” of the southern Appalachians.

There are some exceptions to the rule that expressions current in large sections of the Atlantic states are more widely used in Texas than those confined to smaller areas. Thus general Southern light-wood “kindling” is infrequent in Texas, whereas North Carolinians tow sack “burlap sack” has general currency. The history of the “thing-meant” obviously has a bearing upon the frequency witl which a term is used, and hence upon its dissemination.

The fact that words confined to different speech areas of the Atlantic States are current side by side in Texas leads to the inference—on purely linguistic grounds—that some of the early settlers, or their forebears, ultimately came from the Southern dialect area (the coastal plain) and others from the South Midland speech area (the Appalachians and the piedmont of the Carolinas). Since the English settlements in Texas lag only one or two decades behind those in the intervening areas—Trans-Appalachia and the Gulf States—the connection with the Eastern “homeland” is very close: Tennessee and Arkansas settlers who moved into Texas had only recently crossed the Appalachians, and the cotton belt in Alabama and Mississippi was not taken up until 1820-30 [Paullin-Wright: plates 76E, F]. Data concerning the provenience of the early and the later Texas settlers, though rather fragmentary, agree with the linguistic evidence [Atwood 1962: 7-10].

FIGURE 19: Texas: Quarter till and quarter to From E. B. Atwood, The Regional Vocabulary of Texas.

Atwood is inclined to attribute greater influence upon the regional vocabulary of Texas to the South Midland dialect than to Coastal Southern, but the evidence is far from conclusive. Until phonological and morphological data become available, the question should in my opinion be kept open.

The author finds no lexical evidence for drawing a linguistic boundary along the Brazos River, which on sociocultural grounds— the concentration of slaves in 1860—had previously been suggested as the probable western limit of the Southern dialect area [Kurath 1949: 37]. Nor is there any clear division between the north central and the southeastern section of the state, although some South Midland words have more extensive currency in the former and some Southern words in the latter [Atwood 1962: 82].

In his analysis of the currency of regional words, Atwood introduced two procedures hitherto not employed in this country or abroad: (1) A statistical determination of the frequency of the regional terms within the total area investigated; and (2) a statistical treatment of trends in word usage as inferred from the dissémination by age groups. These calculations were carried out with the aid of an electronic computer.

As to trends in usage, it is important to remember that many of the regional words brought in from the East fall within the semantic domain of farm and home life. Their currency has declined sharply in the nineteenth century, what with urbanization, merchandising of food stuffs, mechanization of the farm, extended schooling, and changes in fashion. Atwood effectively chronicles their decline generation by generation and convincingly points out specific sociocultural factors that underly their recession. His findings are significant not only for Texas and its neighbors; they are symptomatic of trends in word usage throughout the United States.

Atwood’s investigation of this aspect of regional vocabulary demonstrates the effectiveness of combining geographic dissémination with age distribution in tracing linguistic change, a method equally applicable to phonological and to morphological data. A further dimension—the social dissemination of the variants—can be added to achieve the full potential of area linguistics in its effort to deal with language in its sociocultural context.

For the Spanish contributions to the vocabulary of Texas and their regional behavior see 4.6 below.

In preparation for the sociocultural interpretation, Atwood outlines the history of the population of Texas. The salient facts are briefly mentioned below.

English settlement began in 1821 on the lower Brazos River and grew so rapidly that by 1835 the Spanish settlers were outnumbered two to one in a total population of about 35,000. The rout of the Mexican military forces in the battle of San Jacinto (near Houston) led to the formation of the Republic of Texas (1836) and its admission to the United States in 1845. During the days of the Republic the population had risen to more than 100,000; by 1861—when Texas joined the Confederacy—it was 600,000, concentrated in the eastern third of the state. As late as 1900 only the eastern half of the state had more than 6 inhabitants per square mile, although the population had increased to about 3,500,000. Between 1900 and 1930 the High Plains of northwest Texas were occupied, largely from the older parts of the state. Meanwhile urban centers grew apace. Whereas in a population of 4,500,000 two-thirds were classified as rural in 1920, by 1960 three-fourths of the 9,500,000 inhabitants of Texas lived in cities.

The expansion of the English settlements from east to west leads to the expectation that linguistic usage became established to some extent in the eastern section by 1861, where the planter class dominated economically and politically to such an extent that Texas joined the Confederacy. This raises the question of the provenience of this social class: Did the majority of the planters come via the cotton belt from the Atlantic coast and the Gulf plain or from the southern Upland via Tennessee?

The increase in population from 1845 to 1860 and again from 1870 onward is so phenomenal that continued influx from the outside clearly overshadows the growth of the indigenous population. Hence the question is to what extent the newcomers adopted the usage of the indigenous families when it differed from theirs.

The rather late development of important cities is a significant fact. They can hardly be expected to have had much influence upon the speech of their rural surroundings. Their phenomenal growth from 1920 onward implies that country folk flocked into them. What happened, and is happening, as a consequence can only be revealed by rather thorough future investigations of the speech of these centers and their hinterland.