Genealogy traps : common errors in genealogical databases

As in many other fields of research, there are 2 main reasons why mistaken facts are first developed in Genealogy - error and fraud. Just reading an old document can be very difficult indeed, let alone interpreting it correctly. Even then one is still left with the need to question how accurate the original document was in the first place. Clerks made errors in the past just as they do today. However, sometimes they were pressured or bribed to find support for fraudulent claims as newly powerful lords wanted to show they had illustrious pedigrees. In England, descent from William the Conqueror was often valued; in Scotland and Ireland, chieftains sometimes wanted to hide their Norman or Fleming ancestry and pretend to be of Pictish origin; whilst, on the Continent, a family tree showing descent from Charlemagne was often prized. As false claims were being made right from the time records were first being kept, it is simply not possible to produce a wide-ranging genealogical database that is completely reliable. Genealogy is as much an Art as it is a Science. When developing a new database, all anyone can do is to be clear in presenting the data, identify sources, and highlight alternatives, preferences & remaining concerns.

Making assumptions can be very tempting but can be very costly. We've come across cases where people have pulled together most interesting family trees, with potted histories provided about some of the people involved, only to find that the key connection (to a family that was crucial to the relevant tree) had been based on similarities of names and coincidental dates - and was wrong. We suggest that, if your family's family tree is based on links to just one or two well-documented families, you should consider how reliable those connections really are before relying on them to form the basis of your family history. Any early disappointment that disproving a connection may cause will be much easier to deal with than not finding out it is wrong until after you have spent many hours following its trail.

The following are reported in no particular order. They will be added to from time to time. Clicking on the §T§ sign next to the name on the relevant page will bring you back to this page.

1. Gundred, wife of William de Warrenne, 1st Earl of Surrey (on Warren01)
Gundred is widely identified as a daughter of William the Conqueror. When we first uploaded data on the Warrens we were using BE1883 (Warren) as our Main Source for the page and that described her as William's daughter so that is what we showed. However, it appears that she was not the Conqueror's blood daughter but was (possibly) his step-daughter. Such would be an easy mistake to make but it makes a huge difference to her ancestry. However, even that connection, which we provisionally still follow in the database, is open to doubt as TCP (Surrey) appears to support a source that, whilst confirming that she was sister of Gerbod, Earl of Chester, throws doubt that their father was Gerbod of St. Omer and yet more doubt on the view that their mother was Matilda of Flanders who later became wife of the Conqueror.

2. John 'Monoculus', father of Eustace FitzJohn of Knaresborough (on Vesey01)
Many databases follow the report in BE1883 (Vesci) that John was of the de Burgh family. TCP dedicates an appendix to 'The Origin of Eustace FitzJohn' (Appendix A in volume XII/2) that concludes otherwise. TCP also suggests that it may have been Eustace who was known as 'Monoculus' rather than John but we keep to the tradition of attributing that nickname to John.

3. Ivo de Taillebois, lord of Kendal (on ZZmisc03)
Many different ancestries are ascribed to Ivo but none has yet convinced us as being worthy of support.

4. 'The Countess Lucy', wife of the above Ivo de Taillebois (on Malet1)
TCP dedicates an appendix to 'The Countess Lucy' (Appendix J in volume VII) and describes her parentage as "one of the unsolved puzzles of genealogy." We show what appears to be the most likely parentage but it is not certain.

5. Edward of Salisbury, Sheriff of Wiltshire (on ZZmisc01)
BE1883 (D'Evereux) identifies Edward as son of Walter De Evereux of Rosmar but TCP (Salisbury) reports that that Walter was "a fictitious person". This looks as though it is a good example of how an early genealogist made an assumption about someone being connected to another family simply because of a similarity in name or nickname. Once recorded in writing, that assumption gained a following but that did not make it valid.

6. MacDuff, Thane of Fife (see Alpin1)
This reputed son of Duff (Dubh, King of Scots), made famous by Shakespeare in the play MacBeth (MacDuff was the one who slew MacBeth, being "from his mother's womb untimely ripped"), is often identified as ancestor of the early Earls of Fife and hence appears as an ancestor of countless families. In fact, it is virtually certain that he never existed. According to TSP (Fife), it was John of Fordun (who lived in the latter half of the 14th century) who invented the character. Shakespeare probably first came across him in Raphael Holinshed's 'Chronicles of England, Scotland and Ireland', first published in 1577. Shakespeare (1564-1616) was not one to worry too much about being accurate when depicting historical characters or events. The same may be said about the directors of many modern 'historical' films.

7. Gilbert Crispin and Gunnora 'd'Aunou (on Temp13)
We still have more research to do on this couple who were possibly grandparents of the Countess Lucy mentioned above. As we initially did, many web sites confuse Gilbert Crispin, Baron de Bec, with Gilbert, Count de Brionne. Furthermore, whilst Gunnora is sometimes confused with Gunnora de Crepon, she is often identified as daughter rather than sister of Fulk d'Aunou.