John S. Kessler(1)

and Donald B. Ball(2)

Paper presented Saturday, May 20, 2000, at Melungeon Heritage Association Third Union convened at University of Virginia's College at Wise, Wise, Virginia.


Abstract. Recent research into the history, origins, and lifeways of the Carmel Indians of Highland County, (south-central) Ohio, has produced the most comprehensive study to date of this little known Melungeon-related settlement since the studies of Berry (1963), Gilbert (1949), Morgan (1946; 1955), and Price (1950a; 1950b). This effort draws upon archival sources, firsthand observations of the group as it existed in the 1940's and early 1950's, and more recent fieldwork. The present comments have been extracted from a more detailed study of this group scheduled to be released by Mercer University Press in late 2000.


Notice: The present summation of research on the Carmel Melungeon settlement of southern Ohio is released and made available with the express permission and authorization of Mercer University Press, Macon, Georgia. This material is copyrighted and may not be reproduced in any manner for other than personal use.


Ladies and gentlemen, it is truly a pleasure to be here today and share with you the results of a portion of our research on one of the lesser known Melungeon settlements. By way of introduction, I am Don Ball, archaeologist with the U.S. Army Corps of Engineers in Louisville, Kentucky, and with me is my collaborator and co-author, John Kessler, who was raised on a farm near Carmel and interacted on an almost daily basis for nearly 20 years with the folks we will be discussing. The present observations represent but an extended extract of the information presented in a book-length study of the Carmel settlement scheduled for release later this year by Mercer University Press in Macon, Georgia.

To the casual tourist and many area residents alike, the countryside surrounding the small, sleepy crossroads settlement of Carmel nestled at the very edge of the Appalachian foothills in Highland County, southern Ohio, may seem an unlikely place to initiate research into an obscure group which originated in the mid-Atlantic seaboard. As is the case with the majority of the estimated 200 such mixed-blood groups recorded throughout the eastern United States, relatively little scholarly attention has been specifically directed to the study of the Carmel Melungeons. The earliest published reference to this group appears to be but a simple, brief mention of its existence in a general guide to the State of Ohio prepared by the Ohio Writer's Program (1940:509). Such historical and ethnographic information as is available appears principally in the studies of Price (1950a; 1950b), Morgan (1946; 1955), Gilbert (1949:426-427), and a scattering of comments in other sources (e.g., Ayers 1971; Berry 1963; 1978; McBride and McBride 1990). Though insightful, none were either intensive or systematic.

As will be discussed in much greater detail herein, it is a working premise of this effort that the settlement commonly known as the "Carmel Indians" is related to, and derived from, the better known Melungeons of southern Appalachia, themselves the subject of some investigation and much speculation since the late 1800's. Tracing the long and winding route traveled by the ancestors of the Carmel natives as they crossed the rugged Appalachian mountains and ultimately came to settle in the Ohio hill country, the present comments will focus on the history, lifeways, and current status of this settlement. The Carmel group has been traditionally viewed as "Indian" by area Whites and, indeed, made claims to Indian inheritance on its own behalf.

For present purposes, it is more than appropriate to clarify the identification of the Carmel enclave as "Melungeon". A number of scholarly and popular writers alike have restricted the area of occupation of the Melungeons to a relatively limited portion of Appalachia generally consisting of Hancock and Hawkins counties, Tennessee, and Lee, Scott, and Wise counties, Virginia. Such a perspective ignores the residency of genetically comparable and similarly named families throughout an area covering at least 29 adjacent counties variously located in northwestern North Carolina, southwestern Virginia, northeastern Tennessee, and southeastern Kentucky. Accordingly, this paper takes the position that the population scattered across this broad area is de facto Melungeon, be they derived from the "core" area of Melungeon occupation or, alternately, from the same source areas in the mid-Atlantic coastal region. It is ill founded to presume that any given family named Gibson or Collins ­ two "core" surnames encountered within both the "classic" Melungeon heartland and Magoffin County, Kentucky, the source area for the Carmel population ­ are necessarily unrelated. Were, for example, one of these families to move from Magoffin County to Hancock County, they would promptly be deemed "Melungeon". Merely living in an outlying county within this region makes them no less so. Simultaneously, it would be erroneous to assume that genetic variation of particular population pockets did not occur within this region. Thus, the Melungeons living along the Tennessee-Virginia border were genetically similar ­ but not identical ­ to those living elsewhere.

Situated at the foot of Long Lick Hill, the small crossroads settlement of Carmel (pronounced "Car'-mul") is located in Brush Creek Township in the southeastern corner of Highland County, (southwestern) Ohio. This hamlet is literally at the edge of the Appalachian escarpment. To the south and east, heavily dissected, forested hill country predominates. To the north and west, the gently rolling topography is more influenced by till plain formations. Hillsboro, the county seat, is approximately 50 miles east of Cincinnati and an equal distance southeast of Dayton. It is situated less than 30 miles due north of the Ohio River.

Never formally platted as a town, Carmel as a community has always been small and rural in nature (Figure 1). Settled as early as 1823 by the holders of land grants for Revolutionary War service, the community was granted a post office in 1856 (closed 1921). A mercantile store was established on the southwestern corner of the crossroads as early as 1870. In addition to the post office, in its "heyday" in the 1890's and turn of the century, this trading center had grown to a population of 80 persons and hosted four retail establishments (grocery and general stores), a resident attorney, two blacksmith shops, a Methodist Church, and a flour mill situated on nearby Rocky Fork Creek.

Figure 1. Carmel environs, Highland and Pike counties, Ohio.

By the 1940's, electricity and telephone service was available during this period but many persons had neither. Water was from individual wells. The store was the social center for the immediate area. Other local attractions were the church on Sunday, the occasional tent revival which used the field next to the school, and election day. The years have witnessed the continuing decline of Carmel's role in the affairs of the adjacent countryside. The settlement's population was estimated to be but 30 persons in 1970 (Ayers 1971:289).

Situated about 0.8 mile south of Carmel along SR 753 was a small settlement referred to locally as Coon's Crossing. This was an aggregation of houses and shanties, many of which were occupied by Melungeons. This small settlement at the proverbial "wide spot in the road" remains little changed with the exception that many early "shanties" have been replaced with used mobile homes.


The first concentration of persons classified as "Mulatto" in Brush Creek Township appear in the 1870 census. As recorded for that year, these residents consisted of six households varying from 3 to 14 individuals in size with a total population of 40 individuals. Three households represented two surnames each. The surnames present at that time (with number of individuals) were: Gipson (17); Jackson (1); Matthews (9); Nichols (5); Perkins (4); Philips (3); and Wairmine (1) (Breakfield 1995:1, 3-5). Both the appearance of multiple surnames within households and the general proximity of these households to one another serve to suggest that these persons represented extended families of related individuals (see Table 1).

The correlation of age and place of birth information as extracted from the census schedules is of particular utility in documenting the appearance of the group in Ohio. As summarized in Table 2, the general migration of the Brush Creek mixed-bloods is clearly shown by virtue of six out of seven of the oldest residents (50+ years) having been born in Tennessee or Virginia (the implications of these states in the history of the group will be discussed below) while 23 of 33 individuals under the age of 50 stated their place of birth as Kentucky. Of the nine persons born in Ohio, seven were under the age of 10 years while only two individuals over the age of 10 were born in that state. The oldest of these, Margaret Gipson, was 21 years old. This information suggests that various families in the group may have experimented with living in other areas of the state as early as 1849 prior to moving to Highland County. As shown by the census schedule, some families had apparently moved to Ohio, returned to Kentucky, and once again decided to move across the Ohio River. Within the cluster of young persons under 10 years born in Ohio, the oldest child was six years old at the time of the census further suggesting the likelihood that the group had settled in Brush Creek Township as early as ca. 1864.

Though speculative and circumstantial, such a settlement date is supported by the events surrounding the Civil War. Southern Ohio was little affected by the comings and goings of clashing armies. Indeed, through the war years the farm economy of Ohio

Table 1. 1870 Census of mixed-bloods in Brush Creek Township,

Highland County, Ohio.


Table 1 (Continued).

* Personal property valued at $200; no data given for other individuals

or families.

Source: Breakfield (1995:1, 3-5).


Table 2. Age/Place of birth correlation among mixed-bloods

residing in Brush Creek Township in 1870.



prospered in response to feeding an ever increasing Union military force (Jones 1962:5). The likely resultant need for farm labor in Brush Creek Township (and other areas) and the point of origin of the mixed-bloods in the mountains of eastern Kentucky, an area not known for strong Confederate sympathies, served to foster an environment which would at least tolerate their movement into the economically marginal hill country.

Their occupancy in the foothills of Highland County simultaneously afforded a desirable juxtaposition of familiar rugged terrain of little economic use to local farmers and access to construction material, food, fuel, and sources of paid employment. Area farmers in turn gained a source of labor which did not demand higher wages and, indeed, may not have sought or desired steady work (Price 1950b:285). Price (1950a:193) estimated the population of this diffuse group in the Carmel vicinity to be approximately 150 in the late 1940's. For the same time period, Beale's examination of the 1950 Federal census schedules revealed a total population of 450 individuals distributed through three counties; specifically, this figure included Champaign County (60; classified as White and Negro), Hardin County (260; classified as White, Indian, and Negro), and Highland County (130; classified as White) (Beale 1957:194).


Though the reasons for their migration specifically to Highland County remain both obscure and conjectural, the roots of the Carmel colony in Magoffin and adjacent parts of Floyd counties, Kentucky, are well established on the basis of both documentary and oral historical evidence (cf. Price 1950a; 1950b). As suggested by available census schedules, marriage records, and interviews conducted by Price in the 1940's, movement between the two areas had long been prompted by a desire to seek employment opportunities north of the Ohio River while maintaining their familial ties to the Kentucky mountains.

Price's examination of applicable census schedules and other records revealed that the ancestors of the Magoffin County (established 1860) group were present in Floyd County (which then included Magoffin County) by 1810. The 1820 Floyd County census listed several of these families as "Free Person[s] of Color" while in the 1850 and subsequent schedules they were variously enumerated as White, Mulatto, and Indian (cf. Table 3). In general, their racial mixture was evidently a matter of long standing and had occurred prior to their entry into Kentucky. As noted in the census schedules, these mountaineers were variously born in Virginia, North Carolina, and Tennessee (Price 1950b:288).

Within Magoffin County (Figure 2), the major concentration of the mixed-blood population was formerly (1940's) reported to reside along Big Lick, a branch of Middle Creek in the eastern part of the county, and nearby portions of Middle Creek, a tributary of the Big Sandy, which extends into adjacent portions of Floyd County (Price 1950a:201; 1950b:286). In the past, other members of the group were reported to live along Mason Creek (a tributary to the Licking River) about 2 miles (3.2 km) south of Salyersville, the

Table 3. Mixed-blood surnames and racial classification

in Magoffin County, 1850-1880.


Year of Census*

* I = Indian; M = Mulatto; W = White.

(1) 1850 reflects Floyd County only (included Magoffin County at

that time).

(2) 1860 and 1870 names reflect Floyd, Magoffin (established 1860),

Pike, and Johnson counties, Kentucky.

(3) 1880 names reflect only Magoffin County.

Note: For some years, various names may represent only single

individuals or family.

Source: Price (1950a:207a-b).


county seat. Various Gibson, Gipson, and Nichols families resided along Mason Creek.

The predominate names encountered along Big Lick were Cole and Perkins. Price observed that:

the Big Lick, in reality a short narrow branch, is the only concentration of them. It contains six houses and some very poor sites for farms, but a map of 1915 showed 16 houses in addition to the school. A former teacher at the school said that it had 68 pupils in 1925, some of them grown, but none advanced beyond the third grade; the school's enrollment of 23 in 1947, over half of whom actually lived on the Big Lick, also indicated the population decrease. This area was dominated by the Cole family and is yet known as the "Cole Nation" (Price 1950b:286-287).

1. 1 John S. Kessler 2388 Cave Creek Road Falls of Rough, Kentucky 40119


2 Donald B. Ball 312 Iowa Avenue Louisville, Kentucky 40208-1427

Part 2