Finding 38George/Josiah Williams

Identifying 38George/Josiah Williams and his relationships has been a challenging research problem. For many years I was unable to find any primary records pertaining to him at all, so much so that I started to think of him as a ghost. Even now he seems to be strangely absent from some records where one would expect to find him. Fortunately, however, the available evidence, when considered as a whole, reliably identifies him and places him among his family.

His Children

The obituary of 19Sallie Thigpen (née Williams) identifies two siblings: "one sister, Mrs. Julia Carter, of Soperton; one brother, Jordan Williams, of Florida." 1

The death certificates of 19Sallie, Julia, and Jordan all name 38George Williams as their father.2-4


38George in 19Sallie's death certificate. Full page.2


38George in Julia's death certificate. Full page.3


38George in Jordan's death certificate. Full page.4

These death certificates insinuate that 38George Williams married 39Lizzie Norris... but no such marriage record seems to exist. So, moving on then, what can we learn about 38George in other records?

Of the three children (Jordan, Julia, and 19Sallie), Julia is apparently the oldest, born 20 April 1877.3 Therefore, we can guess that 38George and 39Elizabeth had married not long before, say 1875. Since people often marry people who lived near them, and since we know of 39Lizzie's entry in the 1870 census,5 let's look at any nearby Williams families in that census to try to find 38George.

Looking for "George" in 1870

Two pages after 39Elizabeth is listed, we find an entry for George Williams, a 19 year-old black male, in a house headed by Sallie Williams, age 40.6



George Williams in the 1870 census. Full pages: 1, 2.6

Although this George seems correct based on his age, location, and name repetition (His mother is Sallie, and our 38George would later have a daughter named Sallie.2), the marriage seems unlikely for the following reasons:

•Social opposition to interracial marriage was firm at that time. Such pressure would have been even stronger for 39Lizzie, the daughter of a respected pastor and Confederate veteran.
•There was a white man named George Williams living near 39Lizzie around the time that they likely married, and thus he would be a more likely candidate—more on this below.
•An Ancestry.com autosomal DNA sample from my father shows very little African-origin DNA: Ancestry.com categorizes it as 1% Cameroon, Congo & Western Bantu Peoples, whereas one might have expected ~6.25% if his 2x-great-grandfather were fully African. Furthermore, records consistently describe Julia, Jordan, and 19Sallie as white, never mulatto.2-4,7-13

Tax Digests

As note above, tax records from Johnson County, Georgia prove that in 1873 a white man named George Williams was living in militia district 1202,14 and district 1203 in 187515 (map of districts). Living in the same districts were John F. Norris (son of 78Jordan Norris16—make mental note of John here because he appears again in an important record later) in 187314 and 78Jordan Norris himself in 1875,15 so we can be reasonably confident that this is our George.


1873: 38George and his future brother-in-law John F. Norris in militia district 1202. Full page.14

1875: 38George in militia district 1203. 78Jordan Norris was in the same district. Full page.15

1873: 38George and his future brother-in-law John F. Norris in militia district 1202. Full page.14


1875: 38George in militia district 1203. 78Jordan Norris was in the same district. Full page.15

Why then can we not find 38George in Johnson County in the 1870 census? Perhaps he moved into the area after 1870? Or was it that his family was simply skipped? (The 1870 census is notorious for having undercounted the southern states, maybe even deliberately so in order to minimize the South's representation in Congress in the aftermath of the Civil War.)

Other than the black Williams family discussed above, there are no other Williams households in 78Jordan Norris' immediate area in the 1870 census. Therefore, let's next look back at the 1860 census to see whether any Williams families were near 78Jordan at that time.

Looking for "George" in 1860 + Newspaper Marriage Announcement

The 1860 census shows three Williams households near 78Jordan Norris' household. The closest is that of 76J. P. Williams, just four households away.17 Other nearby Williams families were Peter's18 (six houses away) and Freeman's19 (eight houses away).

As you can see, none of these Williams boys is named George, although a few match 38George's likely age range (i.e., Josiah, John, Henry, Jesse, and Limon)... Could one of them actually be our 38George, albeit with a different/misidentified name? With that question in mind, now consider the following newspaper marriage announcement, published 16 January 1874 in The Sandersville Herald,20 one county over from Johnson County. (Note that Johnson County's own pre-1900 newspapers were all destroyed in a courthouse fire.)

The pastor S. M. Norris and groomsman John F. Norris are sons of 78Jordan Norris,16,21 and thus brothers to 39Elizabeth Norris. This helps to affirm her identity in the marriage announcement.20

So 38Josiah Williams married 39Elizabeth Norris on 8 January 1874,20 yet 38George Williams and Elizabeth Norris are the parents of Julia, born 20 April 1877.3 ... Are George and Josiah one and the same?

Evaluating (i.e., Disproving) the Hypothesis that George ≠ Josiah

As an alternative hypothesis, one could argue that 38Josiah may have died soon after his 1874 marriage, and that 38George named in the death certificates2-4 is a different man altogether. However, this hypothesis seems exceedingly unlikely on its face (It's unlikely that 38Josiah would have died so young, but even if he had, his death would be remarkable enough that one might expect to find his death mentioned in nearby newspapers, yet I find no such record.), but further consideration of the tax digests casts additional doubt on this alternative hypothesis:

•The previously mentioned 1873 tax digest14 shows that 38George lived in the same militia district as John F. Norris, one of the groomsmen named in the marriage announcement the following year.20
•W. N. Kight, the groomsman and friend who paid for the marriage announcement,20 was also in the same militia district.22
•Admittedly, the tax digests don't directly tell us the taxpayers' ages, but they do hint at 38George's age: 38George/Josiah doesn't appear in Johnson County's 1872 tax digest at all (I manually checked all militia districts.), but then he appears in the 1873 digest,14 as noted previously. Since the poll tax was assessed for people who were ages 21-60, 38George's sudden appearance in the 1873 digest is precisely what you'd expect if he were born circa 1852, as we know 38Josiah was.17 The tax digests also show that 38George owned no real estate or other taxable property, which is quite common for young men who are just beginning their adult lives.
•"Geo. Williams" appears in the 1873 tax digest in militia district 1202,14 and "George Williams" appears in the 1875 digest in district 120315 ... but he's missing from the 1874 digest. However, "J. Williams" (an abbreviation for 38Josiah Williams, perhaps) does appear in the 1874 digest in district 1202. The strength of this evidence is somewhat weakened, however, since "J." might be John (son of 76J. P. Williams17) or Jesse (son of Peter Williams18), both of whom seem to be just barely under the 21 year-old age limit for the poll tax.


1873: Williams taxpayers in militia district 1202, Johnson County. Full page.14

1874: Williams taxpayers in militia district 1202, Johnson County. Full page.23

1873: Williams taxpayers in militia district 1202, Johnson County. Full page.14


1874: Williams taxpayers in militia district 1202, Johnson County. Full page.23

DNA Evidence: Y-DNA

Thus the paper-trail evidence firmly supports the hypothesis that 38George and 38Josiah are the same person, a son of 76Joseph P. Williams, but DNA evidence supports it further:

Don Charles Williams, a male-line descendant of 76Joseph P. Williams via his son John, matches 36 out of 37 Y-DNA STR's with Pleas Williams,24 a (proposed) male-line descendant of 76Joseph P. Williams via his son 38George/Josiah. I've made sketches of their lineages with links to supporting evidence: Don's, Pleas'. As you can see, Don and Pleas are 3rd cousins once removed, and thus there are five generations back to their proposed most recent common ancestor (MRCA), 76Joseph P. Williams. Using Moses Walker's MRCA calculator and assuming a 0.0043 mutation rate, we thus see that there's a 46.5% likelihood that the MRCA was in 76Joseph's generation or closer.

DNA Evidence: Autosomal DNA

Autosomal DNA supports my hypothesis, as well: An Ancestry.com autosomal DNA sample from my father shares 38 centimorgans (cM) on two segments with a sample from Johnny Williams,25 a descendant of 76Joseph P. Williams via two different pathways—one line via 76Joseph's son John, and another line via his son Nathan. I've prepared a brief outline of Johnny's lineage, including links to some relevant evidence.

Johnny and my father are (proposed) 3rd cousins once removed, but multiple instances of pedigree collapse complicate the usual analysis of their shared centimorgans. BanyanDNA simulations show that 38 cM is 0.5 standard deviation below the expected average for such a relationship (average 62, common range 23-120). You can see screenshots of the proposed pedigree and BanyanDNA results here.

Conclusions

Although complicated by large gaps and the George/Josiah alias, the paper-trail evidence certainly supports the hypothesis that 38George (as named in the death certificates of Jordan, Julia, and 19Sallie) is one and the same as 38Josiah named in the 1860 census, son of 76Joseph P. Williams and 77Elizabeth Stokes.* Indeed, "George" and "Josiah" overlap in their geography, their estimated/apparent ages, and in their closeness to relevant family members. The newspaper marriage announcement, when considered in light of the later death certificates, is particularly strong evidence.

DNA evidence—both Y-DNA and autosomal—greatly reinforces my hypothesis, thereby eliminating whatever lingering doubt may have arisen from the gaps in the paper-trail.

Other People with Similar Names

A man named George Williams, born circa 1850 and the husband of Adeline, appears in Johnson County in the 1900 census, but his children's names don't match what we know about our 38George, and he was in Alabama in 1870.

Two women named Elizabeth appear in some Johnson County records, including tax digests and the 1900 census (1, 2), but they are a generation older.

Another man Josiah Williams lived in Johnson County, but he was a generation younger, born 1879. He frequently appears as "Joe" or even "Joseph" in several records (1880 census, 1900 census, WWI draft card), but his death certificate and his grandmother's will prove that his real name was Josiah. Some hasty genealogies that I've seen conflate this Josiah with our 38Josiah, but simply comparing our 38Josiah's estimated birth year as reported in the 1860 census (i.e., 1852)17 versus the death certificate of this other Josiah easily demonstrates that they're not the same person.

Partial Bibliography

I searched for 38George in the following Johnson County records, to little avail:

•
Guardian bonds (1859-1883) index
•
Letters of guardianship book A (1859-1919) index
 
There is a J. B. Williams on page 208, but this doesn't seem relevant.
•
Inventories & appraisements book A (1859-1887) index
•
Inventories & appraisements book B (1887-1941) index
•
Wills book A (1859-1923) index
 
Page 95, the will of Mrs. Jane Williams, dated 1888, mentions her father Shadrick Williams Sr., dec'd, a sister Lourie Daley, and Leticia Williams (relationship not immediately clear) who had three children: Louisa, Benjamin, and Josiah Williams.
 
Page 141, the will of James T. Williams, dated 1900, mentions "M. W. Williams' children." It also makes clear that James' father is still living, but he's not named. Adjoining property owers are named and include M. W. Williams and James Copeland.
 
Page 182, the will of Samuel Williams, dated 1915, mentions wife Holland; land in militia district 1326; daughters Delitha Holland, Sindy Powell (who had a daughter Farist Powell), Matilda Powell, and Martha Dixon; sons T. J. Williams and James W. Williams; and others.
•
Mattie Lee Hoover Meadows' book Searching for Our Ancestors among the Gravestones: A Cemetery Record of Johnson County, Georgia
 
Pages 222-223 cover two Williams cemeteries, apparently including some of the people named in other records above, but none seems relevant for our 38George.
•
Johnson County Centennial, 1858-1958: May 11-16, 1958 Souvenir Program.
 
This book has no index, so I simply perused it somewhat quickly. I found no relevant Williams mentions.
•
Rebecca P. Lovett and Mary Jo Stephens, eds., Bicentennial Memory Book: Johnson County, Georgia
 
This book's call number at the Georgia Archives is F 292 .J6 B52.
 
On page 5 in a section about Adrian titled "Adrian Boomed at Turn of Century with Coming of Railroad, Various Entrepreneurs," the text reads in part, "The Baptist Church was established September 25, 1891, with Rev. J.A. Stephens and Rev. J.T. Chipley with the following nine charger [sic] members: J.F. Williams, Arcena Williams, Joseph B. Williams, Fanny M. Williams, J.M. Edwards and wife, Morgan Johnson and wife, and Mary J. Cox."

Footnote:

*Since 76Joseph P. Williams married 77Elizabeth Stokes in January 1851,26 since 38Josiah was born circa 1852,17 and since 76Joseph and 77Elizabeth were still married in 1860,17 one can easily deduce that 38Josiah's mother is 77Elizabeth Stokes.

Sources Cited:

1: "Mrs. Thigpen Rites Today," The Dublin Courier-Herald (Dublin, Georgia), Wednesday 11 March 1953, page 8

2: Laurens County, Georgia death certificate 6778 for Sally Thigpen. You can see a copy of the certificate here.

3: Treutlen County, Georgia death certificate 7301 for Julia Carter, who died 3 March 1956. You can see a copy of the certificate here.

4: Taylor County, Florida death certificate (state file #9100) for Jerdon J. Williams, who died 25 March 1954. You can see a copy of the death certificate (minus the cause of death) here.

5: 1870 U.S. Federal Census (Population Schedule). Militia District 1202, Johnson County, Georgia. Page 179B, dwelling 99, Jordan Norris household. NARA microfilm publication M593, roll 160. FamilySearch, <https://familysearch.org/pal:/MM9.3.1/TH-266-12151-89953-5?cc=1438024> accessed 18 December 2014.

6: 1870 U.S. Federal Census (Population Schedule). Militia District 1202, Johnson County, Georgia. Pages 180A and 180B, dwelling 112, Sallie Williams household. NARA microfilm publication M593, roll 160. FamilySearch, <https://familysearch.org/pal:/MM9.3.1/TH-266-12151-90687-35?cc=1438024&wc=92KX-N77:518665401,518856501,519583201> (page 180A) and <https://familysearch.org/pal:/MM9.3.1/TH-266-12151-90172-95?cc=1438024&wc=92KX-N77:518665401,518856501,519583201> (page 180B), accessed 18 December 2014.

7: 1910 U.S. Federal Census (Population Schedule). Hickory Head district, Brooks County, Georgia. Enumeration district 20, sheet 13B, line 51, an entry for Jordon J. Williams. NARA microfilm publication T624, roll 174. FamilySearch, <https://www.familysearch.org/ark:/61903/3:1:33S7-9RJ5-9QQX?i=23>.

8: 1920 U.S. Federal Census (Population Schedule). Hickory Head district, Brooks County, Georgia. Enumeration district 45, sheet 3A, line 11, an entry for Jordon J. Williams. NARA microfilm publication T625, roll 237. FamilySearch, <https://www.familysearch.org/ark:/61903/3:1:33SQ-GR67-D7X?i=4>.

9: 1910 U.S. Federal Census (Population Schedule). Oconee, Militia District 1338, Laurens County, Georgia. Enumeration district 110, sheet 2B, dwelling 37, family 37, Wm Thigpen Jr household. NARA microfilm publication T624, roll 199. Internet Archive, <http://archive.org/stream/13thcensus1910po199unit#page/n337/mode/1up>, accessed 10 October 2021.

10: 1920 U.S. Federal Census (Population Schedule). Orland District, Treutlen County, Georgia. Enumeration district 126, sheet 9A, family 105, Willie M. Thigpen household. NARA microfilm publication T625, roll 280. Ancestry.com, accessed 17 July 2010. Internet Archive, <http://archive.org/stream/14thcensusofpopu280unit#page/n1121/mode/1up>, accessed 10 October 2021.

11: 1930 U.S. Federal Census (Population Schedule). Rockledge, Militia District 1720, Laurens County, Georgia. Enumeration district 88-32, sheet 10A, dwelling 184, family 190, Willie Thigpen household. NARA microfilm publication T626, roll 373. Internet Archive, <http://archive.org/stream/georgiacensus00reel373#page/n185/mode/1up>, accessed 10 October 2021.

12: 1940 U.S. Federal Census (Population Schedule). Rockledge, Militia District 1720, Laurens County, Georgia. Enumeration district 87-35, sheet 5A, household 92, W. M. Thigpen household. National Archives, <1940census.archives.gov>, accessed 2 April 2012. This image is also available on FamilySearch at <https://familysearch.org/pal:/MM9.3.1/TH-1971-27876-13994-72?cc=2000219&wc=MRB8-1JY:790105601,795238401,798691501,798691502>.

13: 1900 U.S. Federal Census (Population Schedule). Lothair, Militia District 1221, Montgomery County, Georgia. Enumeration district 78, sheet 16A, dwelling 287, family 289, Charles Carter household. NARA microfilm publication T623, roll 213. FamilySearch, <https://familysearch.org/pal:/MM9.3.1/TH-266-11100-179466-58?cc=1325221>, accessed 18 December 2014.

14: Johnson County, Georgia. Tax digest book for the years 1873-1881, section for the year 1873, militia district 1202, entries for Geo. Williams and John F. Norris. Ancestry.com ("Georgia, U.S., Property Tax Digests, 1793-1892" / Johnson / 1873-1881 / images 30 and 31 of 788).

15: Johnson County, Georgia. Tax digest book for the years 1873-1881, section for the year 1875, militia district 1203, entries for Jordan Norris and George Williams. Ancestry.com ("Georgia, U.S., Property Tax Digests, 1793-1892" / Johnson / 1873-1881 / images 196, 197, 198, and 199 of 788), accessed 2 June 2024. 38George Williams is on image 198, and 78Jordan Norris on 196.)

16: 1910 U.S. Federal Census (Population Schedule). Lothair, Montomgery County, Georgia. Enumeration district 138, sheet 7B, dwelling 12, family 12, Jorden Norris household. NARA microfilm publication T624, roll 201. FamilySearch, <https://www.familysearch.org/ark:/61903/3:1:33S7-9RJZ-45S?i=1>.

17: 1860 U.S. Federal Census (Population Schedule). Johnson County, Georgia. Page 40 or 484, dwelling 269, family 252, J. P. Williams household. NARA microfilm publication M653, roll 128. FamilySearch, <https://www.familysearch.org/ark:/61903/3:1:33S7-9YBT-7GK?i=39&cc=1473181>.

18: 1860 U.S. Federal Census (Population Schedule). Johnson County, Georgia. Page 38 or 482, dwelling 257, family 242, Peter Williams household. NARA microfilm publication M653, roll 128. FamilySearch, <https://www.familysearch.org/ark:/61903/3:1:33SQ-GYBT-W61?i=37&cc=1473181>.

19: 1860 U.S. Federal Census (Population Schedule). Johnson County, Georgia. Page 38 or 482, dwelling 255, family 240, Freeman Williams household. NARA microfilm publication M653, roll 128. FamilySearch, <https://www.familysearch.org/ark:/61903/3:1:33SQ-GYBT-W61?i=37&cc=1473181>.

20: The Sandersville Herald (Sandersville, Georgia), 16 January 1874, page 2, column 5, near the bottom of the page. Georgia Historic Newspapers, <https://gahistoricnewspapers.galileo.usg.edu/lccn/sn85034106/1874-01-16/ed-1/seq-2/>.

21: "DEATH OF AN AGED MINISTER: Rev. Norris, Well Known Throughout Section," The Montgomery Monitor (Montgomery, Georgia), Thursday 26 August 1915, page 1. Georgia Historic Newspapers, <https://gahistoricnewspapers.galileo.usg.edu/lccn/sn89053231/1915-08-26/ed-1/seq-1/>. This record proves that S. M. Norris was a brother to 39Elizabeth Norris.

22: Johnson County, Georgia. Tax digest book for the years 1873-1881, section for the year 1873, militia district 1202, entry for W. N. Kite. Ancestry.com ("Georgia, U.S., Property Tax Digests, 1793-1892" / Johnson / 1873-1881 / image 24 of 788).

23: Johnson County, Georgia. Tax digest book for the years 1873-1881, section for the year 1874, militia district 1202, entry for J. Williams. Ancestry.com ("Georgia, U.S., Property Tax Digests, 1793-1892" / Johnson / 1873-1881 / images 108 and 109 of 788).

24: For privacy reasons, I won't share further details of the DNA match here. This encrypted .7z file contains those details.

25: For privacy reasons, I won't share further details of the DNA match here. This encrypted .7z file contains those details.

26: Laurens County, Georgia. Marriages book G (1809-1855), page 270. FamilySearch, <https://www.familysearch.org/ark:/61903/3:1:33S7-8BZF-LFL?i=488&cc=1927197>.