Irish Origins

If you're Math-phobic, skip this article.  It will explain and demonstrate a proposed method to assist in locating Irish ancestors.

A valuable resource used to trace families back to Ireland is Sir Robert Matheson's Surnames in Ireland (1909). One of my first posts showed how common (or rare) some of the Irish surnames in our family tree were in Ireland in the 1800s, when all of our known Irish ancestors came to the United States. The bulk of Matheson's report is a table showing the number of births registered for every surname (family name) in Ireland in 1890, and the distribution of these births by province (Leinster = eastern Ireland, Munster = southwestern Ireland, Ulster = northern Ireland, Connaught = mid-western Ireland).  Some of the important things to know about this index are: (1) related names are combined and reported as the most common name; (2) 1890 is after the Great Famine (aprox. 1845 to 1852) and deaths and the enormous exodus of emigrants from Ireland in the mid and late 1800s had decimated the population (Population was growing very rapidly before the Famine, peaking somewhere around 8 million people, but was down to about 3.5 million at the time of Matheson's data in 1891), so this data may not accurately portray the distribution of families in the early 1800s; and (3) rare family names, for which less than 5 births were registered throughout Ireland, are not included. In spite of the limitations, because of the sparsity of census-like information in Ireland, this is a valuable resource.

I have used this book from time to time to give me a general idea of where a branch of my Irish ancestors came from. Because comprehensive searching of data has not been easy (at least in the past), I have not actually found any of my ancestors using this data.  But I hope to.

It has occurred to me that this information can be used mathematically to narrow a search for an ancestor. The listings in Matheson's table are essentially the probability of finding a family with this surname in the various provinces.  Using Hogans as an example:

Surname    Births in:        Ireland              Leinster             Munster            Ulster            Connaught
Hogan                              193                    59                     115                   5                     14

can be recalculated as

Surname    Probability of birth in:  Ireland      Leinster         Munster            Ulster            Connaught
Hogan                                          100%          31%              60%                 3%                   7%

[Because of rounding, numbers don't add to 100.] So I would expect that my Hogan family was most likely from southwest Ireland, but may also have been from eastern Ireland. It is unlikely they came from northern or western Ireland.

I recently found a marriage record that Mrs. Hogan's maiden name was Rice. Matheson's data for Rice is:

Surname    Births in:      Ireland              Leinster             Munster            Ulster            Connaught
Rice                                 99                     33                     18                   48                     0

and can be recalculated as

Surname    Probability of birth in:   Ireland        Leinster       Munster         Ulster         Connaught
Rice                                               100%           33%            18%             49%               0%

Nearly half of the Rice families were in northern Ireland, where there weren't many Hogans, but there were many in eastern Ireland and several in southwestern Ireland. By combining this data, multiplying the probabilities that both families were present in the province, and normalizing:

Surnames    Probability of marriage in:   Ireland      Leinster     Munster     Ulster      Connaught
Hogan-Rice                                           100%         46%           49%         6%              0%

Note that this method assumes that the bride and groom were actually from the province in which they were married. In this case, a Hogan and Rice married in Ireland were likely from eastern or southwestern Ireland, only slightly different from the conclusion I would have drawn from considering Hogan alone.

Applying this to the other families in our family tree for which I know both surnames:

                   Probability                                                                                                        Start in
Surnames   of marriage in:   Ireland      Leinster      Munster       Ulster       Connaught      counties:
Hogan-Rice                        100%          46%           49%            6%              0%             Dublin

Donnelly-Larkin                  100%          38%            7%           50%              5%       Dublin, Armagh

Cushing-Casey                   100%            6%           87%            0%               8%       Cork, Limerick

Casey-Brady                      100%          60%           12%          20%               8%            Dublin

Shannon-McHugh              100%            5%            2%            56%             37%             ?
Waters-Murphy                  100%          24%          76%             0%              0%          Wexford
Murphy-Stafford                 100%         78%             9%           13%             0%       Wexford, Dublin

Unfortunately, I don't think the underlying data for this table still exists.  If it did, we could further analyze this data by county.  In Matheson's table, he also indicates in which counties the most births occured.  Using these (unquantified) indicators, I estimated the most likely counties in my table, above.  The only test I have on this method is that the Cushing-Casey family is known to come from Co. Limerick, near where it joins Cos. Tipperary and Cork. The table above tells me that the family was very likely from Munster province (correct), and the county notes would have sent me to Cork and Limerick counties.

Unfortunately, at this time, there are very few couples in my tree that were married in Ireland and for whom I know the wife's maiden name. In the table above, only three are direct ancestors, and for the one most strongly placed (Cushing-Casey) we already know where they're from.  The other four families are parents of in-laws in my tree to help others connect to our family, but not of enough interest to search their origins.