upload image

Who were the Birth Parents of Walter Cowhig

Privacy Level: Open (White)
Date: [unknown] [unknown]
Location: [unknown]
Surnames/tags: Cowhig Edison
Profile manager: Scott McClain private message [send private message]
This page has been accessed 201 times.

Contents

WHO WERE THE BIRTH PARENTS OF WALTER COWHIG?

INTRODUCTION

As described in more detail in his profile, Walter Wellington Cowhig was born in Boston, Massachusetts, on 8 May 1908 and abandoned to the overseers of the poor in Cambridge shortly after his birth. He was adopted in 1912 by William and Ellen Cowhig, Irish immigrants who raised him in Boston. The identity of his biological parents remained a family mystery for decades, but the advent of widespread consumer autosomal DNA testing has made it possible to use the tools of genetic genealogy to explore new evidence of his origins. The purpose of this Free Space page is to summarize that new DNA evidence.

BACKGROUND

ETHNIC ORIGIN OF WALTER’S BIRTH PARENTS

Three of Walter’s daughters have submitted samples for autosomal DNA testing at 23&Me and/or AncestryDNA, and the results for two of them have been uploaded to MyHeritage and GEDMatch. The ethnicity estimates they received from 23&Me, AncestryDNA, and MyHeritage all revealed a consistent pattern: roughly 50% of their ethnic ancestry was identified as likely of Irish origin; roughly 25% as Norwegian/Scandinavian; and roughly 25% as Ashkenazi Jewish.

Walter Cowhig married Mary Constance Dolan, and she was the mother of all three daughters who provided these DNA samples. Mary Dolan’s ancestry is well-documented for at least three generations: all eight of her great-grandparents lived in Ireland. Thus, the ethnicity results from these testing services strongly suggest that the Irish ancestry of Walter’s daughters is attributable to their maternal line, and that the Norwegian/Scandinavian and Jewish ancestry is a clue to the identity of Walter’s birth parents.

With the advent of AncestryDNA’s “Ethnicity inheritance” feature in 2022, the results were presented even more clearly for the one daughter who tested on that platform:

  • the ethnic origin of “Parent 1” was identified as 46% Ireland and 4% Scotland
  • the ethnic origin of “Parent 2” was identified as follows:
    • Jewish peoples of Europe – 30%
    • Norway – 14% and Sweden/Denmark – 6%

Thus, this project began based on the assumption that one of Walter’s birth parents was of Norwegian/Scandinavian origin, and the other was of Ashkenazi Jewish origin.

BASIC METHODOLOGY

The clear ethnicity division has made it relatively easy to distinguish the Cowhig daughters’ DNA matches that are related to their maternal Irish ancestors from those related to their paternal Norwegian/Scandinavian or Ashkenazi Jewish ancestors, i.e., the birth parents of Walter Cowhig. It also makes it relatively easy to group the matches related to Walter’s birth parents into two discrete groups: those related to his Ashkenazi Jewish birth parent, and those related to his Norwegian/Scandinavian birth parent.

Based on traditional documentary genealogical research, I have tried to identify the most recent common ancestors (MCRAs) for as large a group of DNA matches as possible, both for the Jewish family group and for the Norwegian/Scandinavian family group. Once I identified a likely MCRA couple for multiple matches within each group, I tried build as complete a family tree as I could to identify as many of the descendants of the MRCA couple as possible. Finally, I have used the What are the Odds (“WATO”) tool provided by DNA Painter to compare the probabilities of various hypotheses for how Walter’s daughters might fit into these two family trees based on the strength of their DNA matches to the proven descendants of those MCRA couples.

This WATO analysis is being refined as I identify additional DNA matches and reliably establish their place in the two family trees based on traditional genealogical research. The conclusions drawn as to the identity of Walter’s parents are subject to revision as that process continues.

LESSONS FROM X-DNA: A NORWEGIAN MOTHER & A JEWISH FATHER

Initially, it was uncertain which of Walter’s birth parents was the Norwegian/Scandinavian, and which was the Jewish. Luckily, however, that issue has been resolved by a 23&Me match between all three of Walter’s daughters and “SM,” one of a Norwegian cluster of matches, who shared between 98-131 cM across 3-5 segments with the testing daughters of Walter Cowhig. One of those shared segments was a single 14.6 cM segment on the X chromosome, shared by both “SM” and all three of the testing Cowhig daughters.

Because “SM” is a member of the Norwegian DNA match cluster and because he shares this segment of X-DNA with all three of Walter’s daughters, those daughters must have inherited this matching segment of DNA on their X chromosome from Walter, and Walter must have inherited it from his Norwegian/Scandinavian birth parent. Because Walter is male, he inherited X-DNA only from his mother; a son does not inherit any X-DNA from his father. Therefore, this matching X-DNA segment establishes that it was Walter’s mother who was the Norwegian/Scandinavian birth parent. It also follows that his Jewish birth parent must be his father.

Thus, this project now focuses on the following two questions:

  1. Who was Walter’s Jewish birth father?
  2. Who was Walter’s Norwegian birth mother?

The following two sections summarize my efforts to answer those questions.

WHO WAS WALTER’S JEWISH FATHER?

SUMMARY

The MRCAs of the identified DNA matches in the Jewish family cluster are מרדכי אהרן (Mordecai Aharon) Eydeson and Rachel Hirshkovitz. In the mid-1800s, this couple lived in the town of Žagarė (זשאַגער or Zhager in Yiddish), in what was then Shavli Uyezd, Kovno, in the Russian Empire (now Lithuania). At least four of their children emigrated from Žagarė to America between 1889-1920 and at least one more child had her own children who did. Twenty-one of the confirmed descendants of this MRCA couple, through all five of their known children, have been identified among the DNA matches to the daughters of Walter Cowhig. Based on a WATO analysis of those matches, it is highly probable that Walter Cowhig’s birth father was their son Harris Edison, who was born in Russia in about 1868, emigrated to Boston in 1889, and lived there until his death in 1931.

DISCUSSION

The three daughters of Walter Cowhig who were included in this analysis are Ann Moln (a WikiTree member), and her sisters Ellen Downey (1951-2020) and “BF”. All three sisters tested at 23&Me, and Ann also tested at AncestryDNA. The 23&Me results for Ann and BF were also uploaded to MyHeritage and GEDMATCH.

The WATO analysis must be done separately for each of these test-takers. Because her testing results were compared on more platforms than either of her sisters, the WATO analysis for Ann is currently based on 18 confirmed matches (excluding three matches who were children of other matches). The analysis for “BF” is currently based on 11 matches, and the analysis for Ellen is currently based on 7 matches. The accuracy of the WATO tool increases with the number of matches, so the analysis based on Ann’s matches is the most reliable. For that reason, the primary analysis in the following sections is based on matches to Ann, followed by a discussion of the parallel analysis based on her sister's matches.

Most Likely Hypothesis: Harris Edison

Based on Ann’s WATO analysis, the hypothesis that Harris Edison is the father of Walter Cowhig is by far the most likely of all possible scenarios using either original or beta WATO probabilities.[1]

Other Possible Hypothesis: Unknown Brother

The next most likely scenario based on the WATO analysis was that Walter’s father was an as-yet undiscovered third son of the MCRA couple who was also living in Boston in 1907. This is much less likely than the Harris hypothesis: Using original probabilities, the WATO tool identified the Harris hypothesis as about 7 times more likely than this “unknown third brother” hypothesis, and using beta probabilities, it was 10 times more likely. However, the DNA matches leave open the possibility that this “unknown third brother” hypothesis might be correct, if there is evidence that such a third brother existed.

So far, no such evidence has come to light. No plausible candidate (i.e., no Russian immigrant with a last name of Edison or something similar, born between 1853-1873) appears in the Massachusetts census records for either 1900 or 1910. No other reference to such an individual has been found, despite extensive genealogical research into the other branches of this family, including several that were living in Boston around the same time. In fact, Harris’s nephew Irving Edison, son of Harris’ brother Abraham, recounted in a 1986 letter that his father and uncle Harris were the only two sons of the Edison family to come to America.[2] Thus, there is presently no reason to believe that a third Edison brother was present in Boston in 1907.

Other Possible Hypothesis: Abraham Edison

Excluding that possibility, the next most likely hypothesis identified by the WATO tool is that Walter’s father was Harris’s brother (and Irving’s father) Abraham. There is a family story handed down among Abraham’s descendants which holds that at some point late in his life, Abraham temporarily left his wife for a younger woman, until he was eventually persuaded by his adult sons to return to their mother.[3] At an earlier stage of this research, when fewer DNA matches had been identified and before employing the WATO analysis, this family story did make it seem likely that Abraham was Walter’s father, especially given the many strong DNA matches between his descendants and Walter’s daughters. On the other hand, Abraham had moved with his family from Boston to Georgia in 1894, and he was living in Atlanta in 1907, casting some doubt on whether he would have been likely to father a child in Boston when Walter was conceived. Still, he had many continuing business and family connections in Boston, and so he was certainly a possible father, especially considering the family stories about his extramarital affair.

However, with many more DNA matches to now include in the analysis, including for the first time a DNA match to a direct descendant of Harris, that no longer appears very likely. Using beta probabilities based on Ann’s current set of matches, the WATO tool calculates that the Harris hypothesis is more than 2000 times more likely than the Abraham hypothesis. Using original probabilities, the tool calculates that it is almost 60,000 times more likely that Harris is the father. These results make it highly unlikely that Abraham was Walter’s father, especially because Abraham was living with his family in Atlanta when Walter was conceived.

Other Possible Hypothesis: Louis Shapiro

The only other plausible scenario not affirmatively excluded by the WATO analysis based on known DNA matches to Ann is that the father was Louis Shapiro, a nephew of Harris and Abraham. Louis emigrated to Boston in 1890, and later moved with his wife and daughter from Boston to Baltimore in about 1907, so it is possible he was still in Boston when Walter was conceived. However, using original probabilities, the WATO tool calculates that it is 146,752 times more likely that Walter’s father was Harris rather than Louis. Using beta probabilities, the WATO tool excludes outright the hypothesis that Louis was the father.

WATO Analysis Based on Ann's Sister's Matches

The WATO analysis based on the matches to “BF” and Ellen are less dramatic, but both still identify the probability that Harris was the father as very likely. For “BF,” he is identified as the most likely hypothesis using either original or beta probabilities. For Ellen, he is identified as the most likely hypothesis using beta probabilities and the second most likely using original probabilities.

One key difference between the WATO analysis based on Ann’s matches and the analyses based on her sisters’ matches is the match between Ann and “BMG,” the only confirmed direct descendant of Harris included in the analysis. BMG was an AncestryDNA match, and Ann was the only sister testing on that platform. Therefore, BMG was not included in the WATO analyses for the other two sisters. The addition of Ann’s match to BMG into her WATO analysis dramatically increased the calculated probability that Harris was Walter’s father.

APPENDIX - DNA MATCH DETAIL

Following is a summary of the currently-identified DNA matches to Walter’s daughters in the Jewish cluster used in the WATO analysis, including the strength of the match, sorted by the child of the MRCA couple from whom they are descended:

Descendants of שׂרה [Sara] (Edison) Shapiro:

Descendants of רישע [Risa/Rose] (Edison) Shapiro:

Descendants of אברהם [Abraham] Edison:

  • “HLF” (grandson of Samuel Bernard Edison) – AncestryDNA, Ann Moln 117 cM
  • “JEF” (grandson of Samuel Bernard Edison) – AncestryDNA, Ann Moln 119 cM
  • “HL3” (grandson of Rose (Edison) Gordon) – 23&Me, Ann Moln 70.2 cM; BF 98.2 cM; ED 141.6 cM
  • Bernard Alan Edison (1928-2018), son of Irving Edison – 23&Me, Ann Moln 202 cM; BF 244.2 cM; Ellen Downey 274.7 cM
  • “CAE” (granddaughter of Simon Edison) – AncestryDNA, Ann Moln, 129 cM

Descendants of ציפה [Tzipa/Celia] (Edison) Daniels:

Descendants of רצבי [Ratzbi/Harris] Edison:

CAVEAT ON ENDOGAMY

One final caveat is necessary for the Jewish cluster analysis. The FAQs for the WATO tool disclaim that “the tool is not meant for endogamous populations. Endogamy may significantly affect the scores, and we don't yet have good probability data to figure out how.” Thus, the WATO analysis must be used with caution in this Ashkenazi Jewish population. However, the average strength of the DNA matches used in the analysis for Ann Moln was 120 cM, with multiple large segments of matching DNA in nearly every match considered. Given the dramatic differences between the calculated probability that Harris Edison was the father and any other plausible hypotheses, it seems reasonable to conclude that the uncertainties related to the possible impact of endogamy did not materially affect the outcome.

WHO WAS WALTER’S NORWEGIAN MOTHER?

The Norwegian family cluster of matches to the daughters of Walter Cowhig is a smaller set than the Jewish cluster. It currently consists of seven matches to Ann for whom a common ancestor has been reliably identified, ranging in strength from 47 cM to 139.9 cM. One of these matches is to the child of another match, so the WATO analysis based on Ann’s matches currently relies upon a sample size of six.

IDENTIFYING THE MRCA OF THE NORWEGIAN FAMILY CLUSTER

All six of these confirmed matches descend from a Norwegian couple who lived and died in Kristiania (Oslo), Norway, between 1867 and 1933: Christian Lauritz Olsen (1871-1920) and Nilsine Natalie Eriksen (1867-1933).

The two oldest children of this couple were Ludvig Rolf Olsen (1891-1948) and Bjarne Gerhard Olsen (1893-1965). Ludvig emigrated from Norway to America in 1906 and Bjarne followed him in 1909. Both settled in the Boston area, where they both married, raised families, and spent the rest of their lives. Five of the six DNA matches to Ann are their descendants: one descendant of Ludvig and four descendants of Bjarne. One of the descendants of Bjarne is “SM,” whose X-DNA match was discussed above.

Christian Lauritz Olsen and Nilsine Natalie Eriksen had at least seven additional children after Ludvig and Bjarne. All were born in Kristiania, Norway, between 1895-1910. It does not appear that any of the younger siblings ever emigrated to America. The youngest child in this family was Else Olsen, who was born in Kristiania in 1910. Else Olsen married Olaf Pederson Slette in Oslo in 1945, and her granddaughter “MBS” (who still resides in Norway) was the sixth DNA match to Ann Moln.

THE LIKELY IDENTITY OF WALTER COWHIG’S MOTHER

Narrowing the Possibilities Based on X-DNA

As noted above, “SM” – a maternal grandson of Bjarne Gerhard Olsen – shares a 14.6 cM segment on the X chromosome with all three testing daughters of Walter Cowhig. Because SM’s DNA connection to the Cowhig sisters is through Bjarne, SM must have inherited this matching X-DNA segment from that grandparent. Because Bjarne is male, he in turn must have inherited that X-DNA segment from his mother, Nilsine Natalie Erickson, because males inherit X-DNA only from their mothers. Therefore, the MCRA of Walter Cowhig and this Norwegian family group must be Nilsine Natalie Ericksen or her ancestors.

WATO Analysis

The WATO tool suggests based on the cluster of six confirmed matches to Ann Moln that it is very unlikely Walter was a direct descendant of Nilsine Natalie Ericksen. However, the WATO tool suggests it is very likely that Walter’s mother was either Nilsine’s sister or her half-sister. Using WATO original probabilities based on Ann’s matches, the half-sister hypothesis is the most likely. Using beta probabilities, the full sister hypothesis is most likely. For the two other daughters of Walter, with much fewer matches to compare, the full sister hypothesis is more likely under either probability set. However, using either probability set and for all three sisters, the WATO tool calculates that these two hypotheses are far more probable than any other plausible scenario.

Excluding the “Full Sister” Hypothesis

Traditional documentary genealogical research appears to exclude the full sister hypothesis, because there does not appear to be any full sister of Nilsine who could plausibly be Walter’s mother.

Nilsine’s parents were Niels Eriksen (1840-bef 1870) and Laura Birgitte Larsen (1835-1910). Nilsine was the third of three daughters born to this couple between 1865-1867. Her father Niels Ericksen died before 1870, when her mother Laura (identified as an "Enkemadame" [widow]), remarried to her second husband in Oslo on 6 November 1870.

Thus, it seems clear that Nilsine had only two full sisters, and neither emigrated to America:

  • The oldest sister was Emilie Nilsdatter, who married Ole Kristofersen Löbben in Kristiania, Norway, in 1886. They raised a large family of eleven children in Hole, Buskerud, Norway, between 1887-1907.
  • The second sister was Josefine Lovise Nilsdatter (1865-1868), who died as a young child in Oslo in 1868.
The Most Likely Hypothesis: Walter’s Mother was Augusta Lovise Larsen (1879-1949)

There is, however, a plausible half-sister candidate. As described in more detail in their profiles, Nilsine’s mother Laura married her second husband, Gulbrand Torkildsen (aka Gulbrand Larsen), in Oslo in 1870. Gulbrand and Laura had five children together between 1870-1879: three sons and two daughters. The older daughter, August Lovise Larsen (1875-1878), died as a young child in Oslo. Their younger daughter, also named Augusta Lovise Larsen, was born in Kristiania one year after her older sister died. It appears highly probable that this second August Lovise Larsen was the birth mother of Walter Cowhig.

As described in more detail in her profile, the younger Augusta Larsen emigrated on her own from Norway to America in 1901 at the age of 22, bound for Boston, Massachusetts. She apparently found work as a servant in Boston from 1901-1905, and then returned to visit her family in Kristiania in about 1905. Then, in 1906, she sailed from Norway to Boston a second time, the ship’s manifest indicating that she was “returning home to Myrtle St 97” in Boston. Her nephew Ludvig Rolf Olsen, whose passage she guaranteed, traveled with her on the same ship to America. Ludvig also gave 97 Myrtle St. in Boston as his destination on the ship’s manifest. The 1908 Boston City Directory lists “Augusta Larsen” as a resident of 157 Charles St. in Boston, which is just a few blocks from 97 Myrtle St. Both addresses are near the waterfront, near the east end of the Longfellow Bridge, just across the Charles River from Cambridge where Walter Cowhig was abandoned to the overseers of the poor in May 1908. They are also just two blocks away from the furniture store at 273 Cambridge Street operated at that time by Harris Edison, and his home around the corner at 23 N. Anderson.

In 1910, Augusta Larsen was still single, keeping a boarding house at 85 Myrtle St in the same waterfront neighborhood. By 1910, her nephew Bjarne Olsen and her brother Oscar Larsen had also emigrated from Norway to Boston; the two of them and Augusta’s nephew Ludvig were all living in Augusta’s boarding house in 1910.

As described in more detail in her profile, Augusta Larsen became a naturalized U.S. citizen in 1915. She lived for the rest of her life in the Boston area, never married, and died in Somerville, Massachusetts in 1949.

No Other Plausible Half-Sister

The documentary record seems to clearly establish that Augusta's mother Laura Birgitte had no other daughter who could have been a potential half-sister of Nilsine and mother of Walter Cowhig consistent with the known DNA matches:

  • There is no evidence that Laura had an additional daughter by a third father before her first marriage to Niels Ericksen in 1863, and if she did, that daughter would have already reached the end of her childbearing years by 1908 when Walter was born in Boston. Laura was age 43 and near the end of her own childbearing years by the time Augusta was born in 1879.
  • Laura and her family appear in the 1875, 1885, 1891, and 1900 censuses in Norway, when Laura was aged 40, 50, 56 and 65 respectively. The younger Augusta, her two living half-sisters Emilie and Nilsine Nathalie, her full sister Augusta who died as a child, and her three older brothers all appear with their parents in one or more of those census records, but there is no other daughter in any of the three households.
  • The births of the known children of Laura and Gulbrand establish that they had a child every 2-3 years from the date of their marriage until Augusta was born in 1879, with no gaps. The Norwegian church records for the parish in which they baptized their known children are available online, and there is no record for the baptism of any other daughter.
  • Augusta lived with her brother Oscar for many years in Boston until his death in 1931, and her two nephews were also living in Augusta's boarding house on Myrtle Street in 1910. Despite these close associations with other members of her family from Norway, there is no record of any sister of Augusta ever living or associating with her in Boston.

Thus, based on the available evidence, it appears highly probable that Augusta Larsen (1879-1949) was the birth mother of Walter Cowhig.

APPENDIX - DNA MATCH DETAIL

Following is a summary of the currently-identified DNA matches to Walter’s daughters in the Norwegian cluster used in the WATO analysis, including the strength of the match, sorted by the child of the MRCA couple from whom they are descended:

Descendants of Ludvig Rolf Olsen:

  • “RCS” (grandson of Ludvig), AncestryDNA, Ann Moln 63 cM

Descendants of Bjarne Gerhard Olsen:

  • “SO” (son of Bjarne), 23&Me, Ann Moln 139.9 cM
  • “JSM” (granddaughter of Bjarne), AncestryDNA, Ann Moln 47 cM
  • “SM” (grandson of Bjarne), 23&Me, Ann Moln 115 cM, Ellen Downey 98 cM, BF 131 cM
  • “JB” (granddaughter of Bjarne), 23&Me, Ann Moln 74.4 cM

Descendants of Else (Olsen) Slette:

  • “MBS” (granddaughter of Else), MyHeritage 127.2 cM, BF 139.7 cM

Endnotes

  1. ↑ The original probabilities used by DNA Painter are based on the 2016 AncestryDNA Matching White Paper. The beta probabilities are based on a 2020 update by AncestryDNA of those probabilities. For more details, see Lea Larkin, "Improving the Odds," The DNA Geek, (https://thednageek.com/improving-the-odds/ : 25 May 2020).
  2. ↑ Letter of Irving Edison to Christopher Freund, 17 February 1986; shared by Christopher Freund on Ancestry.com; image here. Note, however, that Irving’s family moved from Boston to Georgia when he was only five years old, and that same letter, written when Irving was 86 years old, indicates that he was apparently not aware that Celia (Edison) Daniels and Rose (Edison) Shapiro were also siblings of his father Abraham. The letter does refer to two sisters of his mother, suggesting he may have incorrectly remembered Celia and Rose as sisters of his mother rather than his father. Thus, it remains possible, but very unlikely, that there was another male sibling in Boston of whom Irving was unaware.
  3. ↑ This story was shared independently in email exchanges between Scott McClain and (1) a grandson of Irving Edison, and (2) a granddaughter of his brother Simon Edison.




Images: 1
Letter from Irving Edison to Christopher Freund
Letter from Irving Edison to Christopher Freund
Collaboration
  • Login to edit this profile and add images.
  • Private Messages: Send a private message to the Profile Manager. (Best when privacy is an issue.)
  • Public Comments: Login to post. (Best for messages specifically directed to those editing this profile. Limit 20 per day.)


Comments

Leave a message for others who see this profile.
There are no comments yet.
Login to post a comment.