Talking Appalachian, page 5
About the same time that claims for an Elizabethan heritage arose, one scholar writing about West Virginia speech suggested “a Scottish influence to some extent,” although he cited only one possible piece of evidence: fronting of the vowel in good and school (making these words sound more like geed and skeel).22 The idea received only occasional, inconsistent mention for three generations thereafter, mustering little evidence and failing to make much headway against the more popular Elizabethan argument. This slow development happened in spite of the fact that research on surnames and traditional ballads was finding connections to Scotland and Ireland and that historians were becoming increasingly aware of the numbers of Scotch-Irish (with ancestry from Ulster, the northern province of Ireland) in certain interior parts of America. Other than Combs’s argument for the predominance of English surnames in eastern Kentucky, only one early study in Appalachia sought to compute ancestry using surnames. Using a sample of 228 surnames he had compiled, John C. Campbell estimated that the founding population of Tennessee had been one-third Scotch-Irish, one-third English, one-seventh German, and the remainder composed of other nationalities.23 Research using surnames has limitations and must make assumptions. One of the biggest problems is that many English-looking surnames are disguised names borne by emigrants who appeared to be from England but were not. For example, Robertson and Anderson were brought to the United States by many Scots and Scotch-Irish (in some cases, these replaced McRoberts and McAndrew, the counterparts based on Scottish Gaelic and originally native to Scotland). By the same token, Müller, Weiss, Schmidt, and other German surnames were brought to America, and often to Appalachia, only to be altered to the English-looking Miller, White, and Smith. Obviously, looking below the surface for hidden ancestry is a skill that genealogists must master; one of them recently recounted to the author how the seemingly Scottish surname McInturff was borne by a German family originally named Machendorf. Thus, high proportions of surnames assigned to an English background are well-nigh inevitable in such calculations.24
As stated, evidence for a significant linguistic influence from the English-speaking Scotch-Irish was slow to accumulate. Horace Kephart, a student of North Carolina mountain people and their culture, wrote in Our Southern Highlanders a century ago: “Since the Appalachian people have a marked Scotch-Irish strain, we would expect their speech to show a strong Scotch influence. So far as vocabulary is concerned, there is really little of it. A few words, caigy (cadgy), coggled, fernent, gin for if, needcessity, trollop, almost exhaust the list of distinct Scotticisms.”25 Combs asserted that “Scotch and Irish survivals are negligible. They occur here and there, but rarely.”26 A substantial Scotch-Irish component was first argued by Cratis Williams: “Appalachian speech was determined by the predominance of the Scotch-Irish in the settlement of the Mountain region prior to and following the American revolution.”27 However, he and others labored under a lack of sources from the British Isles other than earlier English literature to use for comparison—no dictionaries, grammars, or descriptions of speech in Ulster; no knowledge of eighteenth-century literature written there; and apparently no personal contacts in Ireland to consult. With respect to Scotland, the situation was somewhat better, but one could hardly use Robert Burns’s poetry and Sir Walter Scott’s fiction very extensively. Such a handicap meant that, almost inevitably, U.S. commentators arrived at the conclusion that mountain speech came from England.
Scholarly bridges across the Atlantic needed to be built, and not surprisingly, this was first done from abroad, by those familiar with the contemporary speech of the north of Ireland on the ground. John Braidwood had already identified thirteen Ulster items he believed contributed to the American vocabulary, including granny ‘midwife’, hap ‘quilt’, and mooley ‘hornless cow’.28 Alan Crozier, the most extensive examiner of vocabulary to his time, documented the Ulster ancestry of thirty-three items in Pennsylvania, including piece ‘distance’, dornick ‘small round stone’, fire-board ‘mantel’, and redd up ‘prepare, tidy up’.29 The Scotch-Irish influence was greatest in western Pennsylvania, he concluded. By far the most extensive consideration of the subject is this writer’s From Ulster to America: The Scotch-Irish Heritage of American English, which is based on fifteen years of research, primarily using local literature and archival materials. It presents in dictionary format nearly 400 terms that can be mainly or exclusively attributed to the Scotch-Irish, with one or more dated citations for each from both sides of the water.30 Comparison of Appalachian vocabulary with Ulster and Scottish sources reveals such connections as the following:
airish
‘chilly, cool’
back
‘to endorse a document, address a letter’
backset
‘a setback or reversal (in health)’
bad man
‘the devil’
barefooted
‘undiluted (for coffee)’
beal
‘suppurate, fester’
biddable
‘obedient, docile’
bonny-clabber
‘curdled sour milk’
brickle
‘brittle’
chancy
‘doubtful, dangerous’
Cohee
‘resident of western Virginia’
contrary
‘to oppose, vex, anger’
creel
‘to twist, wrench, give way’
diamond
‘town plaza’
discomfit
‘to inconvenience’
hippin
‘diaper’
ill
‘bad-tempered’
let on
‘to pretend’
nicker
‘whinny’
One of the more intriguing Ulster contributions to American English is cracker for a white southerner, a term now used especially for a resident of rural Georgia or northern Florida.
Survey Account I: The Linguistic Atlas
The Linguistic Atlas of the United States and Canada is a systematic survey of pronunciation and traditional vocabulary launched in 1929. Interviews were first conducted in New England and then in the Atlantic states as far south as northeastern Florida in the 1930s. Inspired by the work of historians and geographers, the project collected data using an 800-item survey of vocabulary, pronunciation, and, to a limited extent, grammar. Its principal goal was to map individual linguistic features and to show their regional patterns.31 Two larger objectives were to outline dialect regions and to correlate regional terms with settlement and migration in North America and ultimately with their earlier history in the British Isles.
Because Appalachia encompasses a vast area from the Northeast to the Deep South, the region would seem to be too large to form a distinct territory in terms of traditional culture or speech. The Linguistic Atlas project findings confirmed this and led to the positing of the “Midland,” a dialect region somewhat smaller than Appalachia stretching westward and southwestward from its cultural and linguistic seed bed of central Pennsylvania, where English was first planted in the region, to the Carolinas. The Linguistic Atlas’s three-way regional division of the North, the Midland, and the South may fly in the face of popular conceptions that rely on a simpler North-South dichotomy, but it is consistent with many linguistic patterns and, just as importantly, with the historical development of the country (the North versus South distinction undoubtedly reflects nineteenth-century sectionalism). “There can be no doubting,” according to Hans Kurath, “that the major speech areas of the Eastern states coincide in the main with settlement areas and that the most prominent speech boundaries run along the seams of these settlement areas.”32 As we will see, these boundaries pertain to features of speech that are uniquely American as well as ones traceable to Europe. Though far from uniform, the Midland region has a number of terms that set it off from both the North and the South. It is subdivided into the North Midland region (encompassing northern West Virginia, northwestern Maryland, and most of Pennsylvania) and the South Midland region (southern West Virginia, western Virginia, and western North and South Carolina). Linguistic Atlas research from the 1950s through the 1970s extended the North Midland boundary across central Ohio, Indiana, and Illinois, and it extended the South Midland boundary across north Georgia and north Alabama, thus including Kentucky and Tennessee in the South Midland. The upland interior of Virginia, the Carolinas, and Georgia (the South Midland, or, as it is known by historians, the Upper South or the backcountry) and the coastal areas of these states (the Lower South) differ broadly in speech and in numerous other ways, such as traditional types of houses and barns, reflecting the fact that much of the interior of Virginia and the Carolinas was settled from the north rather than from the east.33 Although the boundary between the South Midland and the South proper was formed in colonial times, it is also valid for vocabulary that is uniquely American (e.g., South Midland chigger versus Lower South red bug), as well as for many other phenomena mapped by geographers, such as the birthplaces of country musicians.34
Here is how the founder of the Linguistic Atlas portrayed the process of forming the Midland:
This far-flung Midland area, settled largely by Pennsylvanians and by their descendants in the Southern uplands, constitutes a separate speech area which is distinct from the Northern area—the New England settlement area—and from the Southern area. Its northern boundaries run in a westwardly direction through the northern counties of Pennsylvania, its southern boundary in a southwestwardly direction through the Blue Ridge and through the Carolina piedmont. The South Midland, to be sure, exhibits a considerable infusion of Southern vocabulary and pronunciations. . . . After 1720 large flocks of Ulster Scots and Palatine Germans arrived on Delaware Bay and spread out into the back country of Philadelphia and then westward to the Alleghenies and the Ohio Valley, and then southward through western Maryland and Virginia to the Carolinas. . . . The influence of the English-speaking Ulster Scots upon the speech of certain sections of Pennsylvania and of the southern upland cannot be doubted, but it is surprisingly intangible. The Dutch and the Germans, who spoke their own language for many generations and passed through a stage of bilingualism before they gave up their native language, have left a much more tangible impress upon the English of their areas of concentration.35
Because it aimed to identify speech patterns that were closest to the settlement period, the Linguistic Atlas project concentrated on surveying older, rural individuals whose language had been less influenced by travel, formal education, and urban life. In West Virginia, for example, 30 of the 111 interviewees were in their seventies, and 13 were in their eighties. When it conducted its interviews in the Atlantic states in the 1930s, the Linguistic Atlas found twenty items of vocabulary or grammar used predominantly in the Midland. In keeping with its interests in traditional vocabulary, most of these pertain to the household and farm life:
bawl
(of a calf)
blinds
‘roller shades’
green bean
grist of corn
‘load of corn’
hay mow
‘barn loft’
hull
‘to shell (beans or peas)’
lamp oil
‘kerosene’
lead horse
‘left-hand horse of a team’
(arm)load
‘armful’
(little) piece
‘short distance’
poke
‘paper bag’
saw buck
‘sawhorse’
skillet
‘frying pan’
snake feeder
‘dragonfly’
sook
‘call to cows’
spouting/spouts
‘gutters’
sugar tree
‘sugar maple’
(quarter) till
(the hour)
want (to get) off
you’ns
‘you [plural]’36
The Linguistic Atlas found six terms used predominantly in the South Midland:
clabber milk
‘curdled sour milk’
fire board
‘mantel’
jacket
‘vest’
milk gap
‘enclosure for milking cows’
ridy-horse
‘see-saw’
sugar orchard
‘grove of sugar maples’37
North Midland terms included the following:
jag
‘part-load of corn’
piece
‘snack’
run
‘small stream’
side meat
‘pork from the side of a hog’
smearcase
‘cottage cheese’38
Terms concentrated in western Pennsylvania included these:
carbon oil
‘kerosene’
closet
‘outhouse’
cruds
‘curdled milk’
doodle
‘haycock’
grinnie
‘chipmunk’
Mapping terms can be fascinating, but it provides only one way to look at variations in speech. Maps of speech present simplified, two-dimensional snapshots of complex, dynamic processes, and they have limitations:
1. Speakers often know (if not use) multiple terms for the same thing; for example, in western Pennsylvania, users of grinnie usually know chipmunk.
2. A speaker may use terms that have the same meaning in different situations and styles, such as when talking to strangers or workmates versus talking with family members or members of a close peer group.
3. There are normally social differences in usage in any geographic area, based on social standing, level of educational attainment, and so on.
4. All languages change constantly, and most individual elements do as well.
5. Many words and usages are in the process of diffusing geographically, moving along networks usually centered in urban areas or dispersing through media influences, regardless of whether their users move along with them.
These qualifications do not nullify the findings of geographically based research, but they do constrain the generalizations that can be made. In fact, the Linguistic Atlas was aware of these limitations and sought to take some of them into account.
