FTDNA Humphrey Surname Project - News

12/23/16 Update:

All members with mtDNA test results are now grouped.   Remaining ungrouped Ystr results are highly SNP tested members who currently have no matches.  Soon I'll define distinct groups for those members

12/21/16 Update:

After several years of sporadic activity due to illness in my immediate family,  I'm actively updating our Humphrey project pages now

I've moved most of the ungrouped mtDNA results into HG groups and I plan to finish this process quickly

The only remaining ungrouped Ystr results now are highly SNP tested members who currently have no matches.

Recently, I discovered some new cousins in my family by invoking the Advanced Search option, especially with the combination of Family Finder & full mtDNA Sequence test results 

04/20/13 Update:

FTDNA has a sale through 4/22/13:  e.g.  YDNA-37 marker test is $119, normally $169.

Scientific American's current issue has an article, "Human Hybrids", by Michael Hammer, a longtime FTDNA associate at U of Arizona.  It's interesting for those of us who focus on human population migration studies, and the eventual, future "merging" of genealogical time-frames with population migration histories.  The article discusses new theories/models for Human Interbreeding, based on genetic analysis as well as fossil analysis.

There's a newer theory/model, "Hybridization", which combines the original Out-Of-Africa (Replacement) model and the Assimilation theory/model, which posits interbreeding between African & non-African populations over a long period of time.  The Hybridization model posits rare interbreeding, as modern humans replaced more archaic human populations.

Recent DNA evidence of gene-sharing between Neanderthals & Homo Sapiens, Denisovan interbreeding in Asia, and some ancient African interbreeding,  led to the new theory/model, Hybridization.

A recent non-human example of mtDNA impact on population studies:  Giant Squid are part of 1 global maternal, interbreeding species, instead of 3, as scientists previously theorized.  mtDNA analysis indicates 1 large family lineage.   The previous theory/model was based on a chemical analysis of their beaks. 



03/27/13 Update:

Serious medical problems in my immediate family greatly diminished my time for FTDNA projects -- hence the long time since my last update.

We now have 105 members, up from 75 in early 2011.   I have all but 3 members' results grouped.

The best grouping criterion is a known family tree with a verified ancestral lineage.  When a member has one & makes it available to FTDNA, I define a group with the most distant known ancestor. Otherwise, I use the most fine-grained Haplogroup ID.

As more SNP's are discovered we can make the groups ever more specific.  I.E., we can identify subgroups that are more recent:   e.g., R1b1a2a1a1b4 is more recent than R1b, and a much more specific, smaller group.

Rib currently enjoys more progress with new SNP's, but I1 is finally beginning to enjoy some progress too.   Ken Nordtvedt's excellent web-pages continue to stay on top of that progress.    http://knordtvedt.home.bresnan.net

I'll soon post more details about SNP's, FTDNA's offerings, and the overall research efforts to merge genealogical time-frames with historical human population migration time-frames.

Mike ...

12/11/10 Update:

We now have 75 members and 20 distinct lineages.  We have continued to merge lineages and find more genealogical ancestors.

FTDNA is offering a new SNP test, L338, which has the potential to robustly divide HG I1 into subclades.

Given the  lack of SNP's to currently subdivide HG I1 into subclades, Ken Nordvedt has developed a model using haplotypes. This can be found at http://knordtvedt.home.bresnan.net .   A calculator to determine your subclade exists at:  http://members.bex.net/jtcullen515/haplotest.htm , courtesy JIm Cullen, Doug McDonald, and Ken.

E.G., when I plug in my 67-marker FTDNA test results, I get I1-AS & I1-AS10 as my most likely (Nordtvedt) subclades:

Y-37 Haplogroup / Sub-Haplogroup or Y-67 Haplo-I Subclade Prediction:

Haplo-I Subclades and probabilities are as follows:

I-M253-AS10 =>32%  I-M253-ASgen =>28%  I-L22-N =>27%  I-M253-ND =>5%  I-M253-P =>4%  I-L22-uN1 =>2%  I-M253-AS1 =>1% 


7/07/09 Update:

FTDNA has extended the discount offer described in our 6/13/09 Update (below).

Additionally, FTDNA is now offering 4 new SNP's, L158-L161, which can potentially further subdivide Haplogroup I2b1. Any of our I2b1 members who might be interested in genetic population ancestry and "tribal" migrations, might find themselves in a smaller aubdivision of I2b1, with some likely migration scenarios.

For a current diagram of the Y-Chromosome I-Tree, see Ken Nordtvedt's web-page at http://knordtvedt.home.bresnan.net . The image is titled, "WarpedFounderTree.ppt" He also has information there about the tribes or clans of I2b1 and some likely migration paths and timelines.

Email me anytime with questions.

6/13/09 Update:

FTDNA is offering a very competitive discount from 6/9/09 to 6/24/09: $119 for 37-marker Ystr + mtDNA. This would be a generous offer for new members.

There are some new developments for our 4 I2b1 members: newly discovered SNP's, S165 & S166, divide I2b1 into 5 different groups:

I2b1a* - English - Isles
I2b1a* - Isles - Scot
I2b1 - Roots
I2b1 - Continental &
I2b1c - Continental

There's a *.ppt image of the "I" Haplogroup at Ken Nordtvedt's website: http://knordtvedt.home.bresnan.net The image is titled, "WarpedFounderTree.ppt"

For more specific information, feel free to email me.

5/7/09 Update:

Our Ystr results table now has 13 family or haplogroup clusters and a total of 52 members tested. The E1b group now has 6 members, and all but one are descendants of Jonas Humphrey of Wendover, Bucks, 1585. We have 62 members, some of whom have tested mtDNA only.

We continue to merge family trees and resolve paternity puzzles as results become available. In some cases this has paved the way to merging family trees and adding several generations to genealogies.

11/16/08 Update: Ystr Results Page now has Family Cluster Tables

New testing and upgrades over the past 18-20 months have yielded sufficient results to further resolve our former Haplogroup (HG) categories into family clusters, and I've updated the Y-results page of our FTDNA Humphrey Project website:


The title of each result table now has the Haplogroup and name of the oldest known male-line Ancestor, e.g., I1 Michael Humphrey , if it's a proven family cluster. Otherwise, as before, the title of each table is the Haplogroup.

We now have 6 family clusters with 21 members total, and 7 Haplogroup clusters with total of 26 members. Four of the 6 are in the R1b Hg; the others are I1 and E1b (previously E3b). We have 4 members whose Ystr test results haven't returned, and we have 4 women who have mtDNA test results, 56 members altogether.

Some members with only 12 markers tested (mostly National Genographic transfers) are left in their respective HG result tables, and others with 12 markers tested are included in a family cluster, when there's clear genealogical evidence. Testing with at least 25-37 markers is sufficient to prove a genetic relationship.

We have some members who match others, all with only 12 markers tested. They might be closely related genetically, but without further testing we cant be sure.

Formerly, we had 2 family clusters, I1 Michael Humphrey and E1b/E3b Jonas Humphrey (5 members each). Now we have 6.

In one sense, the E1b/E3b Jonas family cluster has been the most successful: a benefit of the DNA testing has been to put members in touch, and resulting collaborations have merged family trees, sometimes extending from 2 generations to as many as 11 generations in the Jonas Humphrey (1587) family.

Some of the R1b family tree merges were described in earlier Updates. Now they are identified in the Y Results table with their own family cluster table and a title with the most distant known Ancestor.

My I1 Michael Humphrey (1620) family tree extends back 10 generations, we've found 2 new cousins from recent testing, and we've also proved a questionable genetic relationship in the late 18th century in Connecticut.

Merging family trees, finding additional generations of ancestry, etc, required a lot of email among groups of us over the past several years. I'm thinking of using the FTDNA Surname Project discussion group capability to facilitate further similar activity in the future. In that forum we could use our actual names in addition to Kit IDs and ancestral names. There are 2 members who do not want to be identified in email discussions by name, and all others say OK, in response to inquiries I made several months ago

For a start, I could copy this Update to the Humphrey forum at FTDNA for more open discussions.


Meanwhile, I'll continue to be in touch individually and in small groups via email and phone, as always.

Happy hunting ...


05/27/08 Update:

Some recent DNA results from Normandy, France strongly support an earlier hypothesis that Y-chromosome HG's R1b & I1(a) were at least as prevelant in Normandy as in other European regions. For reasons dating back to WWII, the French government has hisotrically restricted publication of French DNA surveys. These restrictions have been relaxed recently, at least in some specific cases.

These results are correlate well with the prevalence of R1b & I1(a) in Scandinavian countries. (I1a has just been re-named I1, due to recent SNP research & changes in nomenclature by the YCC Consortium, FTDNA, and others)

The new data supports the verbal lore of some of our genealogies, mine included, that some of our British male ancestors were from Normandy, possibly from Viking or the Norman invasions.

However, as usual, this is not genetic proof, since R1b & I1(a) are also prevelant in Anglo-Saxon areas in England, preceding Viking & Norman invasions in England. The Franks most likely had a lot of R1b & I1(a) also, predating Vikings, and are a potential non-Viking source of this DNA in Normandy.


05/25/08 Update:

Recent SNP discoveries have prompted the YCC Consortium to upgrade the Y-chromosome haplogroup clades, renaming many of the clades. FTDNA has made corresponding changes to the clade names, and we see these results in our Yresults pages and in the Yresults in our Humphrey surname website at FTDNA.

Here's a summary of those changes:

R1b & R1a are unchanged

E3b is now E1b (all our members are verified as E1b1b1, a specific subclade)
E1a is now E1b1a
I2b is now I1b or I1 (depends on the exact SNPs)
I1b is now I2a
I1a is now I1

One of our R1b members had a deep clade test and is now R1b1b2

The complete set of changes (renaming of haplogroup clades) can be seen on the FTDNA home-page by clicking the News icon:

5/05/08 - Family Tree DNA completes 2008 Phylogenetic Tree nomenclature updates. See where you now fit into the tree!

I'll soon add more detail to this message in the "News" page of our Humphrey surname project website at FTDNA.

07/07/07 Update:

We now have 48 members in our FTDNA Humphrey project, the latest a member of the National Genographic Society with 12-marker DNA results, in the R1b1 haplogroup.

Recent searches in Dorset have yielded records of relationships, possibly genetic, with Humphry families in Honiton, Devon, some of whom reflect an earlier surname, Hue(s). We are currently researching the possibilities that Michael Humphrey, 1577, Honiton, was of a former Hues family. There were many Humphrey, alias Hues, families in the Honiton area, well into the 19th century.

We're looking for Hue(s) families for DNA testing, especially any who may be in the Honiton, Devon area, or men who have ancestors from the area - also seeking Humphrey men in that area, and more generally, southern England.

5/05/07 -

Some very recent E3b data: A paper in 2007 by Cruciani, et al, confirms the sub-group, V13 of E3b (also known as E-M78), and estimates the migration into Europe from the Balkans no earlier than 5.3k years ago. This is a large reduction from Oppenheimer's estimate of ~ 10k years ago. Another current designation for V13 is E3b1a2.

Historically, this moves the E3b migration from the farming/Neolithic era up more recently to the Bronze Age. The likley path was from the Balkans into the Mediterranean and up into more nothern Europe.

Thanks to Larry Jones for pointing me to the more recent Cruciani data. Larry sponsors one of our members: Larry is E3b from Llangefni, 20 miles from Abergele, where a concentration of E3b have been tested in Wales. Our member, kit#74795, knows his ancestry back to: Griffith Humphrey, b.1792, Llangwnnadl, Lleyn peninsula, Caernarvon, Wales.

More detailed E3b SNP tests are available from FTDNA & also EthnoAncestry.com. It's natural to expect our E3b men with European ancestry to be V13. Larry Jones knows several men with other surnames who are getting more detailed E3b tests.

5/03/07 -

This update corresponds to the recent E3b Bulk Email. For those who have agreed, we're using names as well as kit# ID in our project emails, but since this is a public website, we still use only Kit# ID & ancestral surnames here.

We now have 4 E3b members, our 4th is kit#86056. Three E3b members match 12/12, and a 4th matches 11/12.

#86056 has a 37 marker result & a confirmed SNP E3b result. We have a another member, #86870, whose results are expected soon. This man has a proven male-line lineage back to Jonas Humphrey, Wendover, Bucks, England, born 1620. If his haplogroup is E3b, we will have established the HG for the 2nd of the 2 earliest Humphrey immigrants to America, the other is Michael Humphrey, Lyme Regis, Dorset, born 1620, my male ancestor.

It would be very interesting to see 25 & 37 marker results for our current E3b members who have 12 marker results. E3b is currently tested to be a very small % of men in the British Isles, concentrated mostly in Wales, the Atlantic coast of the isles, and a scatttering in southern England. The % of our project who are E3b is much greater than the % in the British Isles.

At the moment we have reasons to suspect that several of our E3b members would have close matches at 37 markers, indicating desendancy from Jonas, or a male-line lineage to close male ancestors of Jonas, in England or Wales. We know the ancestry of Jonas back to 1567 in Wendover, Buckinghamshire.

In addition to using names, I'm making email addresses available among our project, so that members can contact others, even those who do not have a close match (using FTDNA's criterion). Personally, I would like to see FTDNA ease up their limits, allowing us to see 33/37 matches.


01/20/07 -

caveatt to the following: one of our members has pointed out that there are authors who disagree with Oppenheimer & cite genetic evidence (from Cruciani)that E3b may have migrated to the British Isles much more recently than Oppenheimer's estimate of ~10k yrs ago, possibly as recently as 6k yrs ago, from the Balkans.

Update on E3b & J2 origins: from Oppenheimer's recent book, "Origins of the British"

We have 2 E3b haplogroup members, kit# 41818 & 53677. Oppenheimer's analysis indicates that both these haplogroups arrived in the British Isles as early as 10K years ago, migrating along the Mediterreanean coasts, up thru western Spain, France, then along the western coast of Britain, into Ireland, and especially Wales. Abergale, Wales has a very high concentration still today of E3b, less so J2. This area was the site of a huge copper mining area, before 8k years ago.

So it's possible that E3b & J2 were very early in the British Isles, especially Wales, along with the majority of R1b.

1/4/07 -

An update based on the recent books of Stephen Oppenheimer & Bryan Sykes, Origins of the British, and Saxons, Vikings, and Celts, resp.:

These authors agree that current genetic data (Capelli, Rootsi, & new data from Sykes) indicate up to 75% of British Isles genetic composition today derives from original post-glacial LGM settlements of R1b men from the Mediterannean, via Iberia & France. All the historical invasions & incursions have made a relatively modest impact on the current genetic landscape in the Isles. The Normans likely had the least genetic impact, tho they had an enormous cultural impact. Ditto the Anglo/Saxons, except for a greater genetic impact likely in eastern England (primarily east Anglia).

Oppenheimer clearly indentifies the R1b Basques of northern Spain & southern France as ancestral to those settlers in the British Isles, likely speaking a language predating Celtic, though genetically, the Celts were descended also from these tribes, not from some "kingdom" in central Europe. The Basques were accompanied by other Mediterraneans & southeastern Europeans, immigrating into the British Isles via western coasts of Iberia & France.

Sykes differs on this point and labels the earlier settlers in the Isles as Celtic, not drawing the distinction of Oppenheimer. Both authors agree that the crucial early post-LGM settlement period was approximately 15k yrs bp to ~ 7k years bp (the beginning of the Neolithic farming era). The essential difference appears to me to be a simple labelling preference.

I welcome opinions. Especially from anyone who reads these books and wants to comment.

Impact on my view of the linguistic origins of our Humphrey(s)/Humphrie(s)/ ... surnames:

Written history suggests Scandinavian origin is most likely, during & shortly after the Hunnish invasions into Europe, 4th century. The popularity seems to have accelerated after 1066 & William's invasion at Hastings.

The recent genetic data & analyses suggest that it's equally or even more likely that the origin was "Briton", and probably was adopted more by R1b men, predominantly Celtic, as defined by Sykes. I'm basing this opinion simply on the "probable" great majority of R1b men in the population in the 4th century. Of course, the authors have their own methods of projecting back in time, but both are using specific haplotypes in their data, just as I have graphed some of it in my personal website. I'm always open to new data & new analysis.

I've long suspected another common opinion of these 2 authors: a significant degree of peaceful settlement in the Isles, as well as the later recorded militaristic ones. It appears that many Viking incursions in the 8th & 9th centuries were launched from the Shetland & Orkney islands, north of Scotland, after these 2 isles were settled peaceably by Norwegians. The YDNA & mtDNA indicate Norwegian married couples settling, not a typical Viking "male takeover".

The Sykes data shows our specific I1a* haplotype in Scotland over England by a factor of 2.

It's interesting to compare this data with that of Oppenheimer who used Capelli & others. In


The graphs use the British Isles data of Capelli, et. al, and show distribution of several of our Humphrey haplotypes in the I1a, R1b, R1a, I1b haplogroups. Soon, I'll be identifying more of our specific haplotypes in these graphs, using our project kit ID#s, in addition to the current I1a lineages.
2006 -

A synthesis of English history, linguistics, and census data combine to suggest the surname Humphrey is characteristically Scandinavian in southeastern England. East & West Sussex are the counties of highest concentration since ~ 1100. Others are Yorkshire, Essex, Norfolk, Suffolk, Kent, & Surrey. English history suggests Norman/Scandinavian ancestry in southeast England, consistent with our (I1a) family verbal lore & tradition of pronouncing "Humphrey" as "Umphrey".

One of our collaborators, though not a member of our FTDNA project, is an Umfreville, who has a verified paper trail to ~1100 in Rouen, Normandy. He is R1b, lives in England today, with many cousins in Canada. French/English history strongly suggests that his paternal ancestors were Scandinavian Viking - Norwegian or Danish.

More generally, in eastern England, like other lineages, our ancestors could be Anglo-Saxon, Danish or Norwegian Viking, or Norman/Scandinavian, likely R1b or I1a. Or possibly pre-Roman era Celtic indigenous folks, likely R1b. Or somewhat less likely, J2, E3b, I1b, I1c.

Elsewhere in England, the spellings Humphreys & Humphries predominate in more western counties, by factors up to 26x (Staffordshire). No surprise, these forms of the surname are very prevalent in Wales & Scotland, indicating more likely Celtic origins, perhaps from pre-Roman times. Haplotype R1b is prevalent in these counties and in our project members who trace ancestry to these counties.

***** This is still a work in progress - comments, contributions, any suggestions/corrections are welcomed. *****

Humphrey (& variations) surname distribution in England - Linguistic/historical/genetic synthesis - Historical timeline throughout 4 eras:

SUMMARY: Data for the 3 most common spellings of Humphrey, Humphreys, & Humphries, are consistent with English history, and the linguistic history of our surname. Humphrey is a more prevalent spelling in the south & east, Humphries/Humphreys is more prevalent in the west. The DNA evidence thus far in our 35 members suggests R1b & I1a as the most prevalent in England: R1b everywhere & I1a in the south, especially southeast. We have also, I1c, I1b, R1a, J2, & E3b members, supporting the generality of our surname in various spellings in England since the 4th century.

(1) The earliest known forms are Hunfridus, Humfridus, Hunfridis, Umfridis, & other phonetic equivalents, in Scandinavia, Britain, & Germanic regions, & in the era of the 4th century Hun invasions in Europe. "Hunfridus" may have been derived from a phrase meaning: "one who is peaceful with the Huns", or, "One who is capable of peace with the Huns", a somewhat ambiguous phrase. Closely related in time & phonetically are: Unfrey, Umfrey, Umfray, (British Isles); Unfroy, Umfroy, or Honfroi/Honfroy(France).

None of the pre - year 1100 names above are found (searched with www.192.com) today in the British Isles, except for 5 occurences of Honfroy. (From Norman French: Onfroi - sounds like "Onfwa"). The Norman version may have been also derived from "Homme Vrai", meaning "true man", or "free man".

These names with a prepended "H" were commonly used as a first name from the earliest records, e.g., Humfrey. These were adopted as surnames many centuries later: e.g., Humfrey in the 16th & 17th centuries in England.

It's likely that some sporadic immigrations into England (& other European countries) occured in the era of Huns, 4th century, extending to the Viking era, early 8th century onward. The easiest "trips" would have been into the northern islands, Orkneys, Shetlands, Faroes, Isle of Man, Scotland, and northern England.

There are only a few religious records with some of these surnames - no other records that are verfiable.

Using Capelli's 2004 paper & DNA results, the most prevalent Y-chromosome haplotypes were R1b, I1, & R1a.

(2) In the 5th, 6th & 7th centuries Anglo-Saxons invaded & settled the British Isles. Many were Germanic/Danish & from other nearby regions. Any of our ancestors who had surnames at that time would likely have had the Scandinavian/Germanic names listed in (1) above & likely not any of the French variations.

The settlement was very likely widespread, but given the closer geographic proximity of Germania to eastern England, it's easy to suspect more A/S immigration there. There is very strong evidence that like Western Ireland today, all the British Isles in this A/S era had a predominant Y-chromosome haplotype of R1b, & invoking Capelli again, the most prevalent A/S haplotypes would have been R1b, I, & R1a. The R1b numbers would have increased, and more I & R1a added.

As in the earlier era, there are few records of Humphreys in this period - those that exist are Humfridus, Unfridus, etc., the Scandinavian & Germanic forms.

(3) The Viking era in the 8th, 9th, 10th, & early 11th centuries saw many Danish Viking invasions & settlement in eastern England, and Norwegian invasions & settlements in NW England, especially in the Shetlands, Orkneys, Isle of Man, & Scotland. The isles today have the highest concentration of R1a, as in Norway, again using the Capelli data from 2004.

Toward the end of this period & in the next, (post-1066), we see the Norman form of our surname, Onfroy, Honfroy, Homfroy, Homfroi, etc in France and England, especially in southeastern England.

Norway is much higher in R1a than other European countries today, so we assume it was so in this era also, and the Northern isles (& Iceland) have a very high degree also: Shetlands & Orkney

Danish, German, & Norwegian Y-chromosome results today have a higher % of "I1" than Great Britain, except for the 2 "hotspots", Norfolk & York. These areas & to a lesser degree, all of East Anglia (today's Norfolk, Sussex, Suffolk, Essex) were alternately invaded/settled by Danish & Norwegian Vikings, might have accumulated percentages of R1b & I from Denmark & Norway, increasing the R1B & I1 populations. We really don't know how much population change may have accumulated from these "ping-pong" invasions. The Capelli genetic data appears to be consistent with history, except possibly for R1a, which is low throughout the British Isles, with the exception noted above in the Shetlands & Orkneys, where it appears to have a strong Norwegian source.

However, if we assume that the Danish Vikings contributed more to the male gene-pool in Norfolk & York than the Norwegians, then we have more consistency, but this raises other questions to follow.

As before, surname records are scant prior to 1066.

(4) Hastings in 1066 and William's Norman era: Humphrey de Tilleul & other Normans were given property by William and British records show various forms of the Norman names Honfroy, Onfroi, etc. evolving as forenames & surnames. The counties adjacent to Hastings in East Sussex are today the strongest concentrations of Humphrey as a surname, and after 1066, the earliest forms were Umfrey, Unfrey, Umfray, Unfray, Omfrie, etc - consistent with no leading letter "H" & a French pronunciation. Humphrey/Humfrey initially appears as a forename, not commonly a surname till the 1600's. Humphreys, Humphries, Humfries, etc appear as surnames very early after 1066, mostly in the western counties of England.

All of our project members with an "s" ending surname are currently R1b.

Observation: Cultural influences likely played a major role in initial selection of surnames, whenever the selection was made. Post-1066 a Norman form of the surname might have been popular in southeastern England, while in western England, a Welsh form might have been more popular - for any haplogroup.

So now let's add a current Humphrey/Humphreys/Humphries surname distribution to the information above & compare with Capelli's genetic data:

Using www.192.com to search for electoral records today in England -

"Humphrey" is most prevalent in Surrey, Kent, York, London, East Sussex, West Sussex, Essex, Norfolk, Hampshire, & Herfortshire. Mostly eastern England.

"Humphreys" is prevalent in the same counties, *except* for East Sussex (where Hastings is located), & Norfolk. "Humphreys" is much more prevalent, up to 26x more, in some western counties, e.g., Staffordshire.

"Humphries" tracks "Humphreys" fairly closely, sometimes greater, sometimes less, but rarely much different.

CONCLUSION: the current census data for the 3 most common spellings of Humphrey, Humphreys, & Humphries, are consistent with English history, and the linguistic history of our surname. Humphrey is a more prevalent spelling in the south & east, Humphries/Humphreys is more prevalent in the west. The DNA evidence thus far in our 33 members suggests R1b & I1a as the most prevalent in England: R1b everywhere & I1a in the south & east, especially southeast. We have also: I1c, I1b, R1a, J2, & E3b, suggesting the generality of our surname in various spellings in England since the 4th century.