In April 1768, one Michel (Miguel) Deville wed Marguerite “Kasteberg” or “Kastzeberg” in St Louis Cathedral, New Orleans, Spanish Louisiana. The sacramental record from the cathedral is heavily damaged, but suggests Michel was from “Port Royal.” His bride was of German and French extraction.
A few months later, Deville was awarded a large land grant—more than 300 acres—in what is now Alexandria, Louisiana. There, he raised a large family while his enslaved workers tended a successful plantation of tobacco, corn, and indigo. Six of his children went on to have descendants of their own, including Winston De Ville, a prominent Louisiana genealogist and author who passed away only last year.
Most Deville genealogists have assumed that Michel’s father was a Swiss Huguenot who joined the unsuccessful 1732 Purrysburg settlement in South Carolina, about 25 miles as the crow flies and a couple days by boat from Port Royal, South Carolina. Others suspect his father was French. No one knows for sure.
A few online trees use a dit name for Michel. Dit names are a French-Canadian custom, like nicknames, but often at the family level. (Dit is a conjugation of the French word dire, meaning “to say,” and is pronounced “dee.”) A dit name could carry as much weight as a surname; some families even adopted their dit names permanently, abandoning their original surnames. Dit names could describe a personal attribute (e.g., dit Belle-Humeur for good-natured) or a place of origin (e.g., dit le Basque).
Although documentation is lacking, the dit name used for Michel Deville in the online trees is Maltaise.
Was Michel Deville Maltese?
A few weeks ago, one of Michel’s descendants reached out to me because she noticed something odd. Kristy and her 92-year-old mother, Barbara, both have a small percentage of Maltese in their AncestryDNA origin estimates. Barbara has 4%, and it’s on her Deville side.
In fact, Barbara has hundreds of AncestryDNA ThruLines for Michel Deville, and the majority of them have a small percentage of Maltese DNA.
Some of her more distant matches are nearly 100% Maltese.
Deville is not a traditional Maltese name, so perhaps Michel’s mother was Maltese and his father French or Swiss. However, another interesting clue is that Barbara has male Deville matches at 23andMe in the I-CTS11871 haplogroup (equivalent to I-BY3797), which is not uncommon in Malta. Was Michel Maltese? If so, was he full or half?
I advised Kristy to seek out a male-line descendant of Michel Deville to do a Y-DNA test at FamilyTreeDNA. There are Maltese men who have tested there, and this test should tell us whether the Louisiana Deville family derives from the same ancestors.
Extrapolating Ethnicity Estimates
Can we gauge whether Michel was fully Maltese versus half using the origin estimates of his descendants? In theory, yes.
We know how much Malta DNA Barbara has, so we can extrapolate back six generations to her 4th great-grandparent Michel, doubling the Maltese estimate every generation. However, Barbara is 4% Maltese, and that would put Michel at 256%, which is obviously impossible. On the other hand, some of Michel’s descendants have 0% Maltese DNA, which, no matter how many times you double it, is still zero.
What we need is an average, ideally across dozens of people. This is where AncestryDNA’s ThruLines shine. Barbara’s ThruLines identify 669 matches who are descended from Michel Deville via all six of his children who had families of their own. To simply things, I only considered the 133 individuals who were also 4th great-grandchildren.
If Michel were 100% Maltese, those 133 descendants would average about 1.6% Maltese today. If he were half, they’d average about 0.8%. The average was 1.7%.
Is this proof that Michel Deville was 100% Maltese? No, of course not, but it’s strong evidence. Ultimately, we’ll want to supplement it with Y-DNA results and, ideally, historical documentation. If no one has been looking at Maltese documents, perhaps that’s why no one has found them.
Another intriguing observation: there are two Maltese men who are in FamilyTreeDNA’s Big Y Time Tree on the I-BY3797/CTS11871 branch. Both have the surname Vella. Did Vella become Deville in Louisiana?
Considerations on this Approach
Some considerations on this approach are:
- Ancestry’s reference panel for Malta origins is robust. It has at least 1,673 people and has excellent statistics for accuracy.1,2 We can assume that the estimates are reasonably close to the truth.
- Very few Maltese people immigrated to colonial Louisiana. The descendants of Michel Deville are unlikely to have had other ancestors who could have contributed Maltese DNA.
- ThruLines automate some of the work in finding descendants of a shared MRCA. However, I did not confirm that the lineages were accurate, so there is room for error there.
- Some of these people might be descended from Michel Deville more than once. That was not accounted for.
- Chromosome mapping would help to confirm that Barbara’s Maltese DNA segments originated with Michel Deville.
- The Y-DNA results will also be extremely helpful in tracing Michel’s paternal lineage.
1 AncestryDNA, 2024, Ancestral Regions 2024 White Paper (accessed 30 Jul 2026). The 2025 White Paper does not specifiy the size of the Malta reference panel. It had 1,673 people in 2024 and is probably larger now.
2 AncestryDNA, 2025, Ancestral Regions 2025 White Paper(accessed 30 Jul 2026)
Dear Geek,
I live in Malta and my partner Nick is 100% Maltese with no known non-Maltese ancestors for at least 6 generations. He is on FTDNA (Kit No. 345171).
Michel DeVille sounds like a French variant of a Maltese original ‘Michael Vella’ – both Michael and Vella being very common names in Malta, and suggests that his original Maltese ancestor might have been one of the many Maltese who emigrated to Algeria in the 18th and 19th centuries. My partner has quite a significant number of matches with French names in his list, which mystified us at first because as I mentioned, he knows of no non-Maltese ancestors in his family tree. But then we found out that the siblings of at least two of his gr. x 4 grandparents emigrated to Algeria and the present-day descendants are now French speakers, if not now actually French nationals. So if Michel DeVille or his father were amongst those Maltese emigrants to Algeria, that might account for Michel DeVille’s French-sounding name but probable Maltese ethnicity.
Another point to remark is that these French/Algerian matches share much more DNA with Nick than expected for 5th – 6th cousins, sometimes over more than one segment, which suggests that there was probably quite a bit of endogamy amongst the Maltese emigrant community in Algeria – not surprising since the local population would have been Muslim, while the Maltese were and still are very devout Catholics. This prevalent endogamy would skew the results and might account for why your extrapolation of Barbara’s 4% came to an impossibly high percentage.
Amongst his autosomal matches Nick has one man called Stephen Vella whose Y haplogroup is the I-BY3797 that you mention, and another two called Vella whose haplogroup is I-BY3799, so it sounds like some of your Vella men are his relatives, which would connect them to a pure Maltese family with French/Algerian diaspora.
Hope this helps with your research. Might be totally off the mark, of course, but it’s an idea to consider.
This is fabulous information! Thank you. I am very eager to see the Y-DNA results as I agree that they’re likely to match the Maltese Vellas at FTDNA.
I put the 6 men called Vella in Nick’s list into FTDNA’s matrix, and they are all related to each other in varying degrees. Only the three of them that I mentioned give their BY haplogroup, but since all 6 are related to each other, some as close as 1st cousin (1006cMs), we can probably assume that they all share the haplogroup of your DeVille male-line descendants.
There are four Vella men in the I-M223 Y-DNA Haplogroup Project. Two have done the Big Y test and the other two are at Y-12 and Y-25, respectively. They list their paternal ancestors as Matteo Vella, born c.1560, Francesco Vella born c.1850, Matteo Vella, c 1560, and Luciano Vella born c.1890. The Y-12 tester doesn’t give a paternal ancestor name.
Nick’s matches are:-
S Vella – I-BY3797
EP Vella – I-BY3799
CJ Vella – I-BY3799
MD Vella – I-FT139418
E Vella – I-M233
AC Vella – no Y Haplogroup given.
All 6 are related to each other.
Thank you for this! I edited out the given names to protect privacy but passed along their full information to Kristy.
“some of Michel’s descendants have 0% Maltese DNA, which, no matter how many times you double it, is still zero” I demand a recount LOL
Seriously though, it’s interesting.
Sometimes I just write things to amuse myself!
Have you considered the Acadian Expulsion? Port Royal in the context of Nova Scotia is where Annapolis Royal sits today.
That was my first thought! This guy’s story doesn’t “read” Acadian, though, for a few reasons. First, Deville was not an Acadian surname. Also, the Acadians preferred to stick together, settling near one another and marrying other Acadians, while Deville married a German and settled far away from the other Acadian settlements.