he most often-repeated claim about the origins of the Gilbert surname in Devonshire might not be true. This claim is that the family took its last name from the two prominent Norman landholders who were known to possess a huge amount of property in the area (according to the Domesday Book and other records around the year 1086). These two nobles, Richard Fitz-Gilbert, 1st Earl of Clare (1035-1090), and Baldwin Fitz-Gilbert, Sheriff of Devon (1022-1090), were the sons of the murdered Gilbert “Crispin” of Brionne and Eu (979-1040). Crispin was, in turn, the son of Count Geoffrey of Brionne and Eu (953-1015) who was the illegitimate son of Richard I, Duke of Normandy. (Incidentally, “Crispin” was a nickname given him due to his hair being so curly as to stand up as the branches of a pine tree.) Crispin’s two sons, Richard and Baldwin, accompanied William the Conqueror in his 1066 conquest of England and were richly rewarded with lands in Devonshire and other areas. However, a meta-analysis I just completed on the descendants of these two men strongly indicates that they were not likely the forefathers of the Gilbert line.
The root source of this erroneous claim is probably a misreading of the manuscript called View of Devon and Compton, by an author named Westcot. It has been cited in many subsequent books and webpages as claiming the Gilberts “possessed lands in Manaton, (in or near Dartmoor,) in Edward the Confessors’ days”. The original passage in the manuscript reads:
“This riveret parts Manaton, alias Magneton, and Lustlegh. Many have possessed lands here: in the Confessor’s time Gilbert; after Sauls, Horton, Le Moyn, and others.”
Clearly, Westcot simply states someone named “Gilbert” possessed lands there. Further, “Edward the Confessor’s days”, which were 1042-1066, were at a time when there were likely no “Gilberts” as a family, surnames not being a Norman practice then.
I looked at around 280 individuals who were known to have descended from Richard and Baldwin Fitz-Gilbert between their time and about 1200, when we see our first William Gilbert of Compton appear on the scene. This accounts for what I estimate to be about 80-90% of all their descendants over seven or eight generations, some descendants being hard to account for, illegitimate, etc. By the year 1200, these people descended into around 145 different family or house names – not a single one being “Gilbert” or even “Fitzgilbert”. This is due not only to the daughters marrying into other houses (accounting for at least 50% of the descendants), but also the Norman naming convention at that time that did not favor transmitting a single ‘family name’ down successive generations. For example, Baldwin Fitz-Gilbert’s children became known as ‘Fitz-Baldwin’. Richards became “Fitz-Richard”. A grandchild named Geoffrey would have children named ‘Fitz-Geoffrey’ and so on. Of course, there were instances of grandchildren receiving the first name of Gilbert from time to time, and whose children would be known as ‘Fitz-Gilbert’, but again that Gilbert name would simply disappear in the next generation. This process over 130 years or so, when people started taking surnames, resulted in the descendants of Richard and Baldwin becoming families and houses as varied as Boplande, Munchensy, Walter, Pendergast, St. Leger, Houghton, and Marshal – not a single Gilbert family.
Of course it is entirely possible that one of the lesser-known descendants, perhaps not being recorded in the tax rolls and deed registers due to little or no inheritance, decided to take great-grandpa Crispin’s name as a surname to honor him. However, this is unlikely in view of what we do know about naming conventions of the time. There were four descendants of the Fitz-Gilbert brothers, who were alive right before or around 1200, who had ‘Gilbert’ incorporated into their name somewhere. Two of these, though, have had their descendants accounted for (the lines terminating through marriage into the houses that became such families as de Neville, de Monmouth, de Briwere, de Wyesham, de Builli, de Bussey, and de Wahull, or becoming other families through the male lines, such as Claire and FitzWalter). One of the others, Gilbert de Vere, Prior of the Knights Hospitaller, Oxford (1122 – ?), son of Avice FitzGilbert de Clare (1091 – 1163), would likely have had three or four generations of descendants before 1200 – likely losing all reference to his first name in the process. The remaining person, Gilbert de Clare de Strigull, 3rd Earl of Pembroke, (1173 – 1185), son of Richard de Clare, 2nd Earl of Pembroke, Tonebridge (1130 – 1176), died in childhood. Considering the 145 descendants I could identify alive around the late 1100’s to 1200, and a likely 15 or so unaccounted for in that same generation, the odds of any being the progenitor of the Gilbert surname are about 29:3 in the wrong direction.
In addition to this, there is good evidence that Gilbert was in use as a surname already by people not found on the Crispin family tree. For one, a Robert Gilbert witnessed the signing of a deed sometime between 1199 and 1216. From Westcot’s manuscript:
“Sciant presentes et futuri, quod ego Willielmus de Vernon comes Devon dedi Deo et Ecclesie Sancti Michaelis de Brumor, &c. Hiis testibus Mabilia Comitissa, Abbate Quarar, Roberto Gerebert, Rich. Cottle, Samp- son Clerico de Plympton, cum multis aliis”
(John Pierce, citing Westcot in his work, The Worthies of Devon, 1810, went on to claim that this Robert Gilbert was actually the Richard Fitz-Gilbert found in the Domesday Book – a mistake placing Richard 100 years in the future!) Another example is one I mentioned in an earlier article, a William Gilbert cited in the Pipe Rolls of Normandy in 1198. These two are further evidence that some Normans not in the Crispin line were using the surname Gilbert already.