Knoxville Merchant Cowans



Smoky Mountain Cowan Tapestry
Cowan Tapestry
Cowan Links

The Tapestry
Families Old Chester Old Augusta Germanna
New River SWVP Cumberland Carolina Cradle
The Smokies


Interpreting a Similarity Matrix
Cowands of Bertie County


Person:John Cowan (71)
Cowans in Alabama 1850 Census, but born in SC


This group contains a number of separate lineages which almost certainly share a common ancestor, though that ancestor can not be identified as such. With few exceptions, descendant YDNA signatures for all members of this group match each other "tightly" and are presumed to have a relatively recent common ancestor. John (71), born in 1735 in Hunterdon County NJ, appears to be the earliest known patriarch within this group. Other members flourished somewhat later than John (71), and are known from records in the Carolina's, Georgia, Alabama, and Tennessee. One interpretation is that descendants, or perhaps kinsmen of John (71), moved south into the Carolinas about the time of the Revolution, and then spread south and west into Georgia, Alabama, Tennessee, and Louisiana. Another interpretation is that John 71 represents an independent importation, and that the other members of the group descend from a common ancestor in England, but represent independent importations.

The exception to this is Kit 273972, whose kit owner traces descent to a John Buchanan who died in Tennessee before 1820. Since the Kit owner is not a "Cowan" per se, there is presumably an NPE event of some description suspected in this lineage. This kit matches the Knoxcville Merchant Group better than anyother kit in the Cowan YDNA Project, but even so, the match is distant, showing up in similarity analysis at no better than than 14.6%. It is unlikely that the kit owner shares a common ancestor with anyone in the Knoxville Merchant group in a Genealogically Meaningful Timeframe. (GMT)

Image:Knoxville Merchant Cowan Group Distribution.jpg

The Knoxville Merchant Group is (some what arbitrarily) named for three Cowan Brothers, Samuel, Nathaniel, and James, who settled in Knoxville about 1792.

Person:Samuel Cowan (18),
Person:Nathaniel Cowan (1),
Person:James Cowan (35)

Samuel and Nathaniel were well known merchants in the Knoxville area, and numerous records of their business dealings survive. Their descendants were prominent members of the Knoxville community. One descendant, Perez Dickinson Cowan, was named for a his mother's family, and was a cousin of the author Emily Dickinson. This line has been linked in some genealogies with Cowans living in the Blount-Sevier County area. The latter lines of Cowans belong to the Seven Brothers Group, with distinctly different YDNA signatures. As a result a male line genealogical connection can be rejected.


A comparison of Cowands of Bertie County vs Knoxville Merchant Group:


The following is a hypothetical lineage showing "A" possible relationship between various descendants in the Knoxville Merchant Group of the Cowan YDNA project. The descendants are not identified as such, but their publically available kit numbers are shown to help understanding of the YDNA similarity data above. NOte that some kits lack suffcieint lineage data to allow placement in this diagram. Note that various ancestors (theoretical and otherwise) are placed at different levels in the diagram based on their likely distance from the unknown Patriarch of the project. In some cases, the connections shown are problematic. For example, the William Cowand line (3 kits) may contain a substantial error in the lineage, as this individual is claimed as their ancestor by others with a different YDNA signature. WIlliam may be the ancestor for some of the descendants in the Knoxville Cowan group, or his presence in their lineage may reflect an error in their genealogy. Current data is insufficient to evaluate this.


The following expands to the extent possible on the descendant publicly provided vita data for each kit in this group. Data is largely being drawn from existing WeRelate articles and on lineages provided in Ancestry's Public Member Trees. Added data marked with *.. Where the provided data is insufficient to identify the person in publicly available sources, nothing has been added.

lineage commonly or sometimes associated with Pequea CreekLineage commonly associated with Cowans of Bertie County
273972John Buchanan, bef 1820 in TennesseeNon surname match. Kit matches this group better than any other in the Cowan project, but with a dissimilarity of 14.6%, its not likely that there is a recent common ancestor with the group.
214148Person:William Cowand (1)c.1725* Bertie, NC* 1768, Bertie NC1760, Bertie NCJohn Cowand*Elizabeth Wimberly Griffin*Elizabeth's maiden name was presumably Wimberly,
and Griffen was the last name of a previous husband
Verification needed.
260128Person:William Cowand (1)c.1725* Bertie, NC* 1768, Bertie NC1760, Bertie NCJohn Cowand*Elizabeth WImberly Griffin*Elizabeths maiden name was presumably Wimberly,
and Griffen was the last name of a previous husband
Verification needed.
9142Person:John Cowan (61)abt 1720 1775 NCMargaret*Ancestry Public Member Trees give 231 family trees that match a John Cowan b 1720 d 1775, married to a Margaret. !45 of those lineages identify him as the son of David Cowan and Mary Fleming of the Pequea Creek line. IThis may indicate a disconnect in lineage's given either for the Pequea Creek Group or for this kit. Researchers in the Pequea Creek Group give different vita for John (d. 1823 in Kentucky). It is not known whether the descendant of this kit identifies John's parents as David and Mary. In any case, the discrepancy points to a lineage problem either for the Pequea Creek Cowans, or for the Knoxville Merchant Cowans, or perhaps just for the majority of Ancestry tree members who identify David and Mary as John's parents.
7379Person:James Cowan (35)1761*1801, Knoxville, TN*1795 Knoxville, TN*Margaret Christian* Russell One of three brothers who were merchants in Knoxville beginning about 1792. The other two were Person:Samuel Cowan (18) and Person:Nathaniel Cowan (1) for whose descendants we do not have YDNA data.
17581John Cowen,1770*1837 Chester Co. SC1788 *or
Robert Cowan*Jane Fleming*Usually Martha McFadden*,
sometimes Agnis Martin*
Ancstry Public Member Trees give about 20 lineages for this couple. Vita are broadly consistent.
291915Robert Cowan. 1750 1800Insufficient lineage data to evaluate this person.
7758Person:David Cowan (32)bc1755 1826, Laurens, SCCecilia Jenny GarrisonAncestry Public Member Trees examined do not usually identify his parents. In some trees they are given as William Cowan and Susannah Fleming, which indicates a confusion with the Pequea Creek line.
246573Person:William Cowand (1)c.1725 Bertie, NC 1768, Bertie NC1760, Bertie NCJohn CowandElizabeth WImberly GriffinElizabeths maiden name was presumably Wimberly,
and Griffen was the last name of a previous husband
Verification needed.
18756Person:John Cowan (70)17701837 Chester Co. SC1788 or
Robert CowanJane FlemingUsually Martha McFadden,
sometimes Agnis Martin
Ancestry Public Member Trees give about 20 lineages for this couple. Vita are broadly consistent, but POB in Ireland sometimes not given, and those that do give it may simply be guessing.
7989Person:Elijah Cowan (2)1800Ancestry Public Member Trees frequently identify his parents as William Cowan and Sarah Stewart. WIlliam is then given as the son of John Cowan and Elizabeth, who settled on Pequea Creek in Lancaster County PA. The Pequea Creek Group has a much different YDNA signature form that of the Knoxville Merchants. Lineage data for the Pequea Creek Group seems accurate, which (if true) would mean that there's a problem with this kits lineage. Need to verify and validate lineage.
9970Person:David Cowan (32)17551826=Celia Garrison Cowan, Laurens County,Cecilia Jenny GarrisonAncestry Public Member Trees do not usually give identify his parents. In some trees they are given as William Cowan and Susannah Fleming, which indicates a confusion with the Pequea Creek line.
26986Person:John Cowan (71)1735 Hunterdon Co, NJ1786 Berkeley WV*Limited information available. Ancestry has numerous entries for John Cowan b. 1735, but they include entries for clearly different people, and few if any can be localized to Hunterdon, NJ. A previous search online identified enough information to create an article, but with low confidence that this is the same lineage. Given the early DOB, this John could be the Patriarchial ancestor of other members of this group, or perhaps represents an independent importation to America. More lineage data from the kit owner is needed to better examine possible relationships with others in this group.
105964James Cowan1822*SC*<19001844,* St Clair, ALMargaret Moore Ancestry Public Member Trees
97821David Cowan 17551826, Antrim, No.Ireland*Probably the same lineage as for kits 7758, and 9970, though there's not really enough consistent information to make the connection. However, this kit has the person dying in Antrim No, Ireland, which does not sound consistent with the death of the wife, Celia Garrison, in Laurens County, [SC]. More information is needed for all three kits. YDNA Dissimilarity indicates that 7758 and 9970 are very tightly related (less than 1% dissimilarity), but 97821 is slightly more distant (about 4%), perhaps indicating that while related not quite so closely. This could be related to the random workings of YDNA inheritance (kit 9970 having accumulated more mutations, or mutations in the fast moving markers), or it could be that while they share a common ancestor, that ancestor is somewhat deeper than David. More lineage data is needed for all three kits.
  • Additional information pending on this line, to be added to the FTDNA lineage data.

Research Needs

  • We know from YDNA studies of the Cowans that descendants of Samuel (18), Nathaniel (1), and James (35) bear similar YDNA signatures to descendants of various Cowans in the Carolinas, VIrginia, in New Jersey. It seems likely that Samuel, Nathaniel, and James came to Knoxville from one or the other of these areas. We do not, however, know how these Cowans, as a group, related to each other.
  • The earliest known ancestral member of this group (John Cowan born 1735) lived in Hunterdon County, NJ, and amply predates the other known ancestral members of the group. The possibility exists that he is either the groups Patriarch, or a close kinsmen. In this interpretation the lineages centered in the Carolionas, Georgia and TN represent migrations from the New Jersey area, probably after the Revolution. Alternatively, John could represent an independent migration and shares an earlier common ancestor with other members of the group who did not migrate.
  • Several members of this group trace descent to William Cowand of Bertie County NC. YDNA data clearly places them in the Knoxville Merchant Group, but there are other persons in the Cowan Project also traceing to this same individual, but which do not share the same YDNA signature, and fall into different haplogroups (R1b1 for the the one group, and R1a1 for the Knoxville Merchant Group. This is borne out in terms of dissimilarity between the YDNA signatures of the two groups which is well in excess of the usual acceptance criteria (13%). This suggests that one or the other (and perhaps both) of these two sets of Cowans have an error in their lineage.