How does the U106 variance in Ireland compare to that in England and elsewhere?

[rms2]

Did you exclude English and Lowland Scots surnames from your variance calculation for Ireland?

No, for these reasons.
1. Most of these haplogroups are far older than the times of paternally inherited surnames
2. Surnames are oftentimes geographically biased, which is what you'd expect but that does not always reflect deep ancestral relationships.

Haplogroups are older than surnames, but surnames are part of history, and a man who has an English or Lowland Scots surname but who lists an ancestor born in Ireland is likely to be of English or Lowland Scots descent in his y-dna. To include such surnames in "Irish" variance calculations is to get something other than Irish variance.

Given the history of Ireland, that should be obvious.

3. Given the NPE rates that ISOGG recommends (http://www.isogg.org/fgd.htm) the odds of a good lineage (zero NPEs) for only 750 years is only 37%. (.96 to the power of @ gens) so odds are the surnames of any specific individual just aren't right all the way back. Statistical analysis, cluster analysis and STR diversity must be considered in context with the surnames.

Even with their obvious limitations, to just disregard surnames in this instance is to get an invalid result.

Once again, Irish history is what it is, and if the argument concerns the possibility that U106 is of ancient provenance in Ireland, then obvious newcomers, like those bearing foreign, non-Irish surnames, must be excluded.

Remember when you were calculating Polish U106 diversity and Robert insisted that all non-Slavic surnames be excluded? I agree, that was probably the right thing to do.

4. It seems like many surnames have at least two or three variant/origin interpretations.
5. Perhaps most importantly, I'm not a surname expert, it is a sensitive topic and we are talking about hundreds of haplotypes - so it takes time.
6. Everyone can do it for themselves by downloading the spreadsheet. I think one key to the Irish, etc discussion is probably Z156*. I list the surnames for that one in reply 142.[/b]

I do appreciate the difficulty of the task. Foreign, especially English and Lowland Scots, settlement in Ireland was so pervasive it is difficult sometimes to disentangle the Gordian Knot of Gaelic versus foreign surnames. Some Gaelic surnames have been anglicized (it is generally known what those are, however).

Just the same, until it is done or at least reasonably attempted, U106 variance calculation cannot be used as any kind of evidence of the possibility of an ancient U106 presence in Ireland. The reason why that is so is obvious: it is lumping foreign U106 - the U106 of historical, non-Irish input - together with native Irish U106.

But perhaps ultimately there really is no such thing as native Irish U106, at least if by "native Irish" we mean Celtic, pre-Viking Era Irish.

[Mikewww]

Haplogroups are older than surnames, but surnames are part of history, and a man who has an English or Lowland Scots surname but who lists an ancestor born in Ireland is likely to be of English or Lowland Scots descent in his y-dna. To include such surnames in "Irish" variance calculations is to get something other than Irish variance.

Surnames may be a part of an individual's personal history/genealogy, but that doesn't mean they are reflective of deep ancestral relationships. The history and prehistory of Celtic ethnicities are much older than the genealogical timeframe.

3. Given the NPE rates that ISOGG recommends (http://www.isogg.org/fgd.htm) the odds of a good lineage (zero NPEs) for only 750 years is only 37%. (.96 to the power of @ gens) so odds are the surnames of any specific individual just aren't right all the way back. Statistical analysis, cluster analysis and STR diversity must be considered in context with the surnames.

Even with their obvious limitations, to just disregard surnames in this instance is to get an invalid result.

I'm not saying you should disregard surnames. I'm just saying we don't have enough analysis of the surnames or haplotypes to adjust for surnames. I don't like subjective measures as it can appear as "cherry picking" data.  I'm not saying anyone else is doing such a thing, but I don't want to.    

Whoever wants to evaluate surnames should do analyses of the appropriate surname projects, geographical locations, haplotype data both of "in" group and "out" group surnames, and then apply statistical theory. I think this is kind of thing that Trinity College has done in the past.  That type methodology is probably very useful.

Remember when you were calculating Polish U106 diversity and Robert insisted that all non-Slavic surnames be excluded? I agree, that was probably the right thing to do.

Yes I remembered and I complied with the request but doesn't mean I agree with surname selection/de-selection or that the identification of Gaelic, Norman, Flemish, Anglo-Saxon, Scottish, etc. surnames and their variants in the Isles is as easy as picking out Slavic surnames.

Just the same, until it is done or at least reasonably attempted, U106 variance calculation cannot be used as any kind of evidence of the possibility of an ancient U106 presence in Ireland. The reason why that is so is obvious: it is lumping foreign U106 - the U106 of historical, non-Irish input - together with native Irish U106.

But perhaps ultimately there really is no such thing as native Irish U106, at least if by "native Irish" we mean Celtic, pre-Viking Era Irish.

You may be right, in fact, I think you are generally right. However, no one that I've seen has done a proper analysis of surnames and haplotypes to determine which Irish are of Anglo-Saxon Invasion Era English descendancy versus something else prior to that.

[Mikewww]

In came the far away stranger
And settled all over the land.
The horse and the cow
The goat and the sow
Fell into the stranger's hand.

The history of foreign (especially English) settlement in Ireland is well known. The source lands of most of those foreigners are far richer in U106 than Ireland is.

I myself have a number of ancestors in my pedigree who were born in Ireland but were very unlikely to have been native Irish. They were Protestants, for one thing, and did not have native Irish surnames, for another.

While variance is important, it cannot be lifted out of the context of known history, haplogroup distribution, and ethnic and linguistic associations.

If variance is the paramount consideration, then a "high variance" figure for U106 in North America must mean that U106 went there "shortly after its emergence".

How do you think I feel? My family's poems are about Cromwell.  As you can guess, we don't feel that good about him. I've got a good several hundred word poem on "The Lament for John MacWalter (Walsh.)"  However, I have some Anglo-Saxon lineages so there is no use in lamenting for and against one's self.

... so I'll leave relating poems to genetics to others, but here is the best I can do on North America U106. I always put everyone with an unknown/unlisted or New World MDKA's into their own category.  As you can guess, there are a lot of them.

Here are the Isles countries to compare with:

Here are the relative variance results for U106 All (subclades)...

Ireland________:  Var=0.96 (N=139)
Scotland_______:  Var=0.96 (N=115)   
Germany________:  Var=0.89 (N=112)   
England________:  Var=0.86 (N=370)   

Here is how the Unknown/New World U106 folks stack up.

Unk/New World__:  Var=0.92 (N=380)

It is an irrelevant number as far as I'm concerned as the USA is obviously a "melting pot" of various cultures....   but LOL, it's hard to beat the Irish anyway.

[Arwunbee]

To date the arrival of U106 in either Ireland or Britain we need to know what shores it occupied on the continent at any given time.  When U106 is treated as a block (which may be a problem) it suggests that ALL U106 west of Poland is RELATIVELY late.

I'm not sure why U106 would get held up in Poland.  The Lichtenstein 1000 BC R1b man who was found in a cave about 100km west of the Elbe, was probably U106.  His haplotype matches five times as many U106 men in ysearch than he does for every one P312.

[gtc]

What is NOT remotely possible is that early migrations to Britain or Ireland were composed of a single, homogenous grouping, a sort of L21 expansion that jackbooted its way, with blue-eyed purity into Ireland. Perhaps that fantasy appeals to some, who knows :)

Methinks this is your real agenda: You can't read the word "germanic" in any context without seeing "nazi".

[Mikewww]

To date the arrival of U106 in either Ireland or Britain we need to know what shores it occupied on the continent at any given time.  When U106 is treated as a block (which may be a problem) it suggests that ALL U106 west of Poland is RELATIVELY late.

I'm not sure why U106 would get held up in Poland.  The Lichtenstein 1000 BC R1b man who was found in a cave about 100km west of the Elbe, was probably U106.  His haplotype matches five times as many U106 men in ysearch than he does for every one P312.

Alan mentioned the Nordwestblock. I'm starting another thread specific to Germanic expansion and alignment with U106.

[whoknows]

Few points on misrepresentation and selective misinterpretation

I have not asserted the concept of 'native Irish R U106' BUT simply aired the scenario that the Haplogroup may have entered Ireland as part of an early admixture, prior to later invasions from England.

My reference to Tim Janzen's findings on the subject were not to promote, justify or argue any personal objective, but included as obviously salient to the subject of the thread: namely a comparison of R U106 variance in Ireland with other locations.

I am not stating that R U106 in Ireland is composed of entirely ancient lineage that pre-dated  Cambro-Norman, Cromwellian or Elizabethan incursion, what I have proposed is line of reasoning in which the Haplogroup, along with others, probably migrated westwards and that it is difficult to conceive why it alone would not cross into the Isles, whiles others happily did so. On that basis I have considered that it may well have entered Ireland at an early date to establish itself.

It is not my intention to advance any personal agenda, my contributions have sought to question an orthodoxy and to argue the case for maintaining an open mind that can at least accommodate the notion of R U106 as having an early  presence in Ireland or concede that we are not able to definitively exclude such as a probability.

Lastly I agree with the reservations expressed regarding surnames, they are are minefield indeed and while offering some helpful glimpses of themselves cannot be held up with any conviction as definitive evidence. In that context it is a vacuous distraction to focus on seemingly English looking surnames to declare, 'hey presto Tim Janzen's finding's are inherently flawed' As was explained what may appear to non Irish names can often turn out to be corrupted variants of Gaelic names or a crude Anglicized version or early Irish Sept Names. Moreover even if all of the names of those assessed by Mr Janzen had genuinely verifiable English names, all that would do is to relocate the consideration and implication of high variance of R U106 in Ireland, to Britain. Any discussion on that would inevitably raise the question, if it had entered the Isles at such a time would it not have had equal opportunity to arrive in Britain's island neighbor?

[Arwunbee]

It is not my intention to advance any personal agenda, my contributions have sought to question an orthodoxy and to argue the case for maintaining an open mind that can at least accommodate the notion of R U106 as having an early  presence in Ireland or concede that we are not able to definitively exclude such as a probability.

I don't think anyone, ever, has said categorically that U106 cannot have had an "early presence" in Ireland.  You're shadow sparring.

[rms2]

. . .

Here is how the Unknown/New World U106 folks stack up.

Unk/New World__:  Var=0.92 (N=380)

It is an irrelevant number as far as I'm concerned as the USA is obviously a "melting pot" of various cultures....   but LOL, it's hard to beat the Irish anyway.

Ireland is a "melting pot" too, not to the degree the USA is, but in the sense that its U106 population is self-evidently mostly the product of outside, non-Irish input and settlement.

The purpose of looking at North American U106 variance is to illustrate an important lesson, i.e., that variance cannot be regarded as anything but a possible upper bound on the age of a haplogroup in a place, and that it must be considered in the light and context of history and the distribution and ethnolinguistic affiliations of that haplogroup.

Otherwise, if variance were the paramount consideration, and history and other considerations are disregarded, we might conclude that U106 arrived in North America "shortly after its emergence".

The same could be said for U106 in Ireland. Remove its variance from other important considerations and one might make some critical errors concerning its presence there.

Surnames, as imperfect as they are, are important clues to one's origin, especially in Ireland. Yes, haplogroups are older than surnames, but haplogroups are older than most languages and cultures and nations, yet those facts of history cannot be ignored. Including surnames that are plainly of non-Irish origin in an Irish variance calculation renders that calculation virtually meaningless, unless one is not really looking for clues to the age of the haplogroup in Ireland.

[rms2]

. . .

Lastly I agree with the reservations expressed regarding surnames, they are are minefield indeed and while offering some helpful glimpses of themselves cannot be held up with any conviction as definitive evidence. In that context it is a vacuous distraction to focus on seemingly English looking surnames to declare, 'hey presto Tim Janzen's finding's are inherently flawed' As was explained what may appear to non Irish names can often turn out to be corrupted variants of Gaelic names or a crude Anglicized version or early Irish Sept Names. Moreover even if all of the names of those assessed by Mr Janzen had genuinely verifiable English names, all that would do is to relocate the consideration and implication of high variance of R U106 in Ireland, to Britain. Any discussion on that would inevitably raise the question, if it had entered the Isles at such a time would it not have had equal opportunity to arrive in Britain's island neighbor?

Reservations or not, the inclusion of non-Irish surnames in the calculation of "Irish" U106 variance renders that calculation invalid for the purpose to which you are putting it, i.e., as some kind of indication of the age of U106 in Ireland. An invalid variance calculation cannot then be transferred over to England.

Look at Mike's U106 variance calculation for North America in his last post. It is higher than most places in Europe. Will you now argue that U106 arrived in North America "shortly after its emergence"? Ready to take on the "orthodoxy" in that case?

Regarding U106 in England or Ireland or anywhere else, it merely provides an upper bound for the possible age of the haplogroup in that place. Barring something strange, like an identifiable bottleneck or proven genetic drift, the haplogroup cannot be any older there than its variance indicates, but it can certainly be a whole lot younger.

Here's an admittedly imperfect analogy. You are a homicide detective working a murder case. Mr. Smith has been found in his office on the 20th floor of the building where he works, slumped over his desk, a bullet wound in the back of his head. You discover that Mr. Smith is 45 years old. Do you immediately conclude that he has been in his office for the last 45 years? Or do you check with the security guards downstairs to find out when Mr. Smith checked in?

BTW, I heard from Tim Janzen regarding his Irish variance calculation. He did not exclude non-Irish surnames before calculating variance:

Tim Janzen
   
10:21 PM (9 hours ago)
      
to me

Dear Rich,

                I only looked at this by country of origin.  I didn’t do anything in regards to trying to sort out surnames.

Sincerely,

Tim

[whoknows]

To equate Ireland as a 'melting pot' with the USA, in terms of scale, intensity and diversity  is asking too much of credulity.

Secondly your claim regarding Ireland that "its U106 population is self-evidently mostly the product of outside, non-Irish input and settlement." is just that an assertion only, it is clearly not demonstrable in terms of empirical evidence, suggested by the use of the term 'evidently' as there is no evidence to determine one way or another. As noted previously we have ultimately only speculation (albeit informed) the rest, including scales of probability, is built upon that foundation. In that factual context we can equally speculate and consider that along with some R U106 arriving via incursions and settlement from England, the Haplogroup had every opportunity to enter Ireland as part of an earlier admixture.

Regarding the observation that: "Reservations or not, the inclusion of non-Irish surnames in the calculation of "Irish" U106 variance renders that calculation invalid for the purpose to which you are putting it, i.e., as some kind of indication of the age of U106 in Ireland. An invalid variance calculation cannot then be transferred over to England."

Such a selective interpretation is semantic sophistry in that  your original criticism to dismiss Tim Janzen's findings was that he used data, which you assert (again you do not know for certain) was drawn from a majority of supposedly English names. That was the contention you presented to claim his results were invalid in terms of R U106 variance in Ireland and the implied conclusions of it arriving at an early stage. If (and that is another significant  area of doubt in the critique) such names were undeniably English then the only aspect of the data that is changed or in any way less  exact is that fact it would then apply to people whose ancestry (according to your evidence free claims) was from Britain. That applying, then the questions concerning R U106 variance would indeed simply be shifted to Britain with the same question and considerations relating to its possible pre 'Germanic' arrival there.

[Mikewww]

To equate Ireland as a 'melting pot' with the USA, in terms of scale, intensity and diversity  is asking too much of credulity....

The size and scope of USA's immigration is different but it is true that Ireland is not monolithic set of genes, even if there is a lot of L21.  Even the legends/myths support the flow of people in (The Book of Invasions.)

I was quite surprised myself to figure out I was really "Welsh-Irish" and not "Old Irish." There are the Normans, Flemings, Belgae, Bretons, Scandinavians and various kinds of Brits/Picts/Scots coming and going.

I have not asserted the concept of 'native Irish R U106' BUT simply aired the scenario that the Haplogroup may have entered Ireland as part of an early admixture, prior to later invasions from England.

My reference to Tim Janzen's findings on the subject were not to promote, justify or argue any personal objective, but included as obviously salient to the subject of the thread: namely a comparison of R U106 variance in Ireland with other locations.

I do remember Tim's post and thought it was odd that he cited Ireland in the way he did.  I don't think there is evidence to support Ireland as the origination point for U106, but rather a pooling point from different sources perhaps.

On the other hand, this does not mean variance calculations, or frequency calculations for that matter, are of little value. They are, they just must be considered in context and weighed with balance.  I'm guessing different folks idea of balance might differ greatly, but a good look from multiple perspectives is usually helpful.

[Mikewww]

Rather than general claims and counter-claims about broad sets of data without additional evidence, why not consider what I've suggested (and what I think Goldenhind would agree with):

When you have a problem, analyze it.  That means break it down into smaller problems, figure them out and then try to re-assemble how everything fits together.  In other words, get down to the cluster level.

You could probably look at some true U106*, but a great place to start might be Z156.

... This paragroup is interesting, at least as it relates to this topic. This is Z381+ Z156+ but L1(null439)-.      Z156xL1 or Z156*:

9 Ireland
7 England
5 Scotland
2 Germany (Baden-Wurttemberg)
1 Wales
1 Low Countries
......
The relative variance of Z156* is fairly old:
Z156*____________:  Var=0.79 (N=26)

This group is predominately 390=24 with 390=25 as well, which are of course off-modal for U106 but match the L11 and P312 modals. There are even some 393=15 guys.
....
Have at it.

fN10310   Bess   zzzUnkOrigin
f147853   Bow   Scotland
f120386   Donald   Scotland, Grampian, Aberdeenshire, Fintray Parish
fN10078   Goll   Germany, Baden-Württemberg
f111387   Gunter   Wales
f165363   Haile   England
f1543   Jarman   England
f175525   Kidder   England
f19095   Minnir   Germany, Baden-Württemberg, Ernsbach
f134007   Pelan   Ireland, Ulster, Co. Antrim, Belfast
fN46336   Pierssens   Belgium, Flemish Region, East Flanders, Belsele
fN35071   Roche   Ireland, Leinster, Co. Wexford, Monart, Ballinure
f117323   Smalley   England, South West, Devonshire, Bideford
f5010   Staple   England
f57352   Stubbs   England, South West, Gloucestershrire, Elmestone
fN32403   Tonckin   England, South West, Cornwall, Camborne
f180037   Westcott   UK
f47991   Wilson   Ireland
f80961   Keddie   Scotland, Fife, Markinch
f90791   MacMullen   Ireland, Leinster, Co. Kilkenny
f33309   McMillan   Scotland
f35043   McMillan   Scotland
f65132   McMillen   Ireland
f93223   McMillen   Ireland
f93708   McMillen   Ireland, Ulster
f94483   McMullan   Ireland, Ulster, Co. Monaghan, Lisgorran
f159837   McMullen   Ireland

I have no dog in this fight and I am clearly no Irish historian or surname expert.

[Jdean]

Mike

Personally I have a lot more faith in surnames than yourself and have argued this point before.

Though I understand your point about compounded % chances of NPEs, if somebody can trace there ancestry back to Ireland we are probably only taking a couple of hundred years max and if a surname is reasonably clearly not Irish then their line likely wouldn't have been there for very long before that (in most cases).

However I also understand your point about not wanting to manipulate the data !!

I wonder though, would it make much difference if you removed the 'IS Ire z unk' group.

From looking through the people in that group it appears a majority of them are in fact brickwalled in the US and possibly put Ireland down as a reasonable guess.

[whoknows]

Mike, you are a voice of reason, as always, and I respect you balanced position on this, along with your obvious knowledge on this subject. I agree with the valid points you raise, although again must emphasize that my discussion here is nothing to do with any personal desire to establish any particular ethnological line. My contributions have been about a principle, in that there are voices asserting as fact, that R U106 is by definition 'Germanic' ,and that as an extension of that claim insist that the Haplogroup's presence in Ireland is due entirely to a later influx of Germanic R U106 via colonization from England.

Clearly not all are equipped with an ability to step beyond orthodoxy to consider different speculations on this matter as being probable, in essence, we are in the presence of a mindset, entirely dismissive of any notion which threatens the cozy reassurances of convention. As you have noted that's not my position and I greatly welcome your perspective as being a timely reminder that there is indeed a middle way based on rational inquiry and a willingness to maintain an open mind.

That said it is clear that a state of polarity has been reached on this thread, I respect the different views held and hope that others may treat my own position with an equal courtesy. On that note of harmony I shall offer no more contributions regarding this particular thread, apart from extending my best wishes to all who have conversed during the past few days.

Sincerely yours 'who knows' with a Haplotype not from Ireland Germany or any place west of Kiev :)

[Bren123]

There exists an assumption that the high amount of R U106 in the UK is due to so-called Dark Age incursions into Britain by Germanic peoples, what is not considered is that this view is derived, ultimately from a biased political propaganda written by Bede, whose writing gave birth to a creation mythology of the English.

His work on this subject has recently received considerable review, in particular the notion of a flood of Angles, Frisians and Jutes arriving in Britain to displace the British 'Celtic' population. What colonization that did take place is now being viewed as happening on a much smaller scale,  a Germanic military and political elite gaining control over areas, presiding over a majority British population,

If that hypothesis were true English would never had become a dominant language!

[alan trowel hands.]

Few points on misrepresentation and selective misinterpretation

I have not asserted the concept of 'native Irish R U106' BUT simply aired the scenario that the Haplogroup may have entered Ireland as part of an early admixture, prior to later invasions from England.

My reference to Tim Janzen's findings on the subject were not to promote, justify or argue any personal objective, but included as obviously salient to the subject of the thread: namely a comparison of R U106 variance in Ireland with other locations.

I am not stating that R U106 in Ireland is composed of entirely ancient lineage that pre-dated  Cambro-Norman, Cromwellian or Elizabethan incursion, what I have proposed is line of reasoning in which the Haplogroup, along with others, probably migrated westwards and that it is difficult to conceive why it alone would not cross into the Isles, whiles others happily did so. On that basis I have considered that it may well have entered Ireland at an early date to establish itself.

It is not my intention to advance any personal agenda, my contributions have sought to question an orthodoxy and to argue the case for maintaining an open mind that can at least accommodate the notion of R U106 as having an early  presence in Ireland or concede that we are not able to definitively exclude such as a probability.

Lastly I agree with the reservations expressed regarding surnames, they are are minefield indeed and while offering some helpful glimpses of themselves cannot be held up with any conviction as definitive evidence. In that context it is a vacuous distraction to focus on seemingly English looking surnames to declare, 'hey presto Tim Janzen's finding's are inherently flawed' As was explained what may appear to non Irish names can often turn out to be corrupted variants of Gaelic names or a crude Anglicized version or early Irish Sept Names. Moreover even if all of the names of those assessed by Mr Janzen had genuinely verifiable English names, all that would do is to relocate the consideration and implication of high variance of R U106 in Ireland, to Britain. Any discussion on that would inevitably raise the question, if it had entered the Isles at such a time would it not have had equal opportunity to arrive in Britain's island neighbor?

What I am curious about is what kind of matches Irish U106 have.  Surely that is the way to go in terms of clearing this up one way or another.   I assume matches thought to be within the last 1000 years or so should tell us something. 

[Mikewww]

MPersonally I have a lot more faith in surnames than yourself and have argued this point before.

Though I understand your point about compounded % chances of NPEs, if somebody can trace there ancestry back to Ireland we are probably only taking a couple of hundred years max and if a surname is reasonably clearly not Irish then their line likely wouldn't have been there for very long before that (in most cases). ...

I have no problem at all with anyone who has a genealogy trail to Ireland calling themselves Irish, but only placed in the modern context of that ethnicity.

What I have a problem with is a 200 year genealogy trail and a surname provided as proof of a deep ancestral relationship with the Iron Age Celts, or Germans, or what have you. There is long period of disconnect from the times of fixed surnames and good genealogical records.

This is where evaluating complete surname projects, high resolution gegraphical sampling in Ireland (or whatever),  and statistical analysis of haplotypes, TMRCAs and surname variants comes in to play.

"Most cases" your 200 year-old surname trail may indicate an Iron Age affiliation, but maybe not. How would we know without a true analysis?

[Heber]

Few points on misrepresentation and selective misinterpretation

I have not asserted the concept of 'native Irish R U106' BUT simply aired the scenario that the Haplogroup may have entered Ireland as part of an early admixture, prior to later invasions from England.

My reference to Tim Janzen's findings on the subject were not to promote, justify or argue any personal objective, but included as obviously salient to the subject of the thread: namely a comparison of R U106 variance in Ireland with other locations.

I am not stating that R U106 in Ireland is composed of entirely ancient lineage that pre-dated  Cambro-Norman, Cromwellian or Elizabethan incursion, what I have proposed is line of reasoning in which the Haplogroup, along with others, probably migrated westwards and that it is difficult to conceive why it alone would not cross into the Isles, whiles others happily did so. On that basis I have considered that it may well have entered Ireland at an early date to establish itself.

It is not my intention to advance any personal agenda, my contributions have sought to question an orthodoxy and to argue the case for maintaining an open mind that can at least accommodate the notion of R U106 as having an early  presence in Ireland or concede that we are not able to definitively exclude such as a probability.

Lastly I agree with the reservations expressed regarding surnames, they are are minefield indeed and while offering some helpful glimpses of themselves cannot be held up with any conviction as definitive evidence. In that context it is a vacuous distraction to focus on seemingly English looking surnames to declare, 'hey presto Tim Janzen's finding's are inherently flawed' As was explained what may appear to non Irish names can often turn out to be corrupted variants of Gaelic names or a crude Anglicized version or early Irish Sept Names. Moreover even if all of the names of those assessed by Mr Janzen had genuinely verifiable English names, all that would do is to relocate the consideration and implication of high variance of R U106 in Ireland, to Britain. Any discussion on that would inevitably raise the question, if it had entered the Isles at such a time would it not have had equal opportunity to arrive in Britain's island neighbor?

What I am curious about is what kind of matches Irish U106 have.  Surely that is the way to go in terms of clearing this up one way or another.   I assume matches thought to be within the last 1000 years or so should tell us something.  

Alan,

You have hit on an interesting point. I don't think anyone have effectively used broad Haplogroup matches to determine a point of origin. I have attempted to take it a step further and plotted my Halpogroup matches, both paternal and maternal, in 23andme, using autosomal Relative Finder and Ancestry Finder. The results are quiet interesting. This is not to be confused with methodologies such as Dienekes Dodecad, which use full autosomal data. I use only the haplogroups from the autosomal matches.

http://m.box.com/view_shared/6np7lxocu4

The vast majority of my matches are paternal haplogroup R with the majority of these L21 and L21*.
The next large group are paternal I and J.
Likewise the vast majority of my matches are maternal haplogroup H with the majority of these being H1 and H1*.
The next large groups are maternal U, J, K and T.
This is consistent with the signatures of the Celtic and pre Celtic settlers in Ireland and is a more powerful confirmation than using single Haplogroup assignment.

Dr Tyrone Bowes uses a similar methodology to link a persons DNA result to a locality and associated clan.

He has produced a series of maps including Surnames, Clan Territories and Castles. Although his sites is commercial in nature, the tools and methodologies used are powerful.

http://www.irishorigenes.com/content/databases
http://www.irishorigenes.com/content/surnames-science
http://www.irishorigenes.com/store/comprehensive-surnames-ireland-map
http://www.irishorigenes.com/store/irish-clan-territories-map

I believe that when we get full Y and MtDNA sequencing, this approach will be validated.

"The idea is quite simple, stick this map on your wall, and stick pins on your Surname and Surnames of the people you match genetically (i.e. your 'Genetic Matches' from Y Chromosome DNA testing) to reveal your GENETIC HOMELAND! The Genetic Homeland is the area where your ancestors lived for 100's if not 1,000's of years surrounded by other Clans with whom he was related, leaving their mark in the place names and DNA of that areas current inhabitants. Each of the approximately 3,600 Surnames on this map have been placed on the area where it clusters based on the 1911 census of Ireland. Each Surname is colour coded based upon its origin (Irish, Norman, English, Viking, or Scottish/Gallowglass)."

Edit: Surnames the science

[Jdean]

MPersonally I have a lot more faith in surnames than yourself and have argued this point before.

Though I understand your point about compounded % chances of NPEs, if somebody can trace there ancestry back to Ireland we are probably only taking a couple of hundred years max and if a surname is reasonably clearly not Irish then their line likely wouldn't have been there for very long before that (in most cases). ...

I have no problem at all with anyone who has a genealogy trail to Ireland calling themselves Irish, but only placed in the modern context of that ethnicity.

What I have a problem with is a 200 year genealogy trail and a surname provided as proof of a deep ancestral relationship with the Iron Age Celts, or Germans, or what have you. There is long period of disconnect from the times of fixed surnames and good genealogical records.

This is where evaluating complete surname projects, high resolution gegraphical sampling in Ireland (or whatever),  and statistical analysis of haplotypes, TMRCAs and surname variants comes in to play.

"Most cases" your 200 year-old surname trail may indicate an Iron Age affiliation, but maybe not. How would we know without a true analysis?

All in all I think we're on the same page here.

BTW in respect to Alan's question, I did look at quite a large no. of Irish U106 haplotypes a year or so ago when this topic was brought up in an extremely similar fashion at DNA-Forums and found most had reasonably close matches outside of Ireland.

BTW2 There is a Norwegian Z18+ tested chap in the Z18 project, hang my head in shame for having forgotten about him.

[GoldenHind]

I don't really want to get caught up in this argument once again.

Since my name has been mentioned with regard to the surname issue, I thought I should clarify my position.

My point about exclusing German surnames from Polish data, which incorporates anyone whose ancestors came from what is currently Poland, was that much of what is now Poland was formerly Germany before 1945. After WWII about a third of eastern Germany was ceeded to Poland, to compensate for the loss of the eastern part of Poland, which was annexed by the Soviets. Someone whose ancestors were Germans from East Prussia and didn't have a drop of Polish blood in them would be listed as Polish because what was formerly East Prussia is now part of Poland. Otto von Bismarck, who was born in the former German province of Pomerania, also now part of Poland, would be classified as Polish. We don't really have the same situation in Ireland.

The main problem I see with looking at Irish surnames is that quite a large number of Irish surnames were anglicized over the centuries. Some surnames in Ireland are undoubtedly English or Scottish in origin, and some are undoubtedly Irish in origin, but I believe there are quite a few which could be either. I have a very Anglo-Saxon surname, but I know there is an Irish surname which has been anglicized into mine. People with that surname in Ireland could have either an Irish or an English origin. It isn't possible to make that determination from the surname alone.

[rms2]

To equate Ireland as a 'melting pot' with the USA, in terms of scale, intensity and diversity  is asking too much of credulity.

No one did that. I even noted that Ireland is not a melting pot to the same extent the USA is, but it certainly has experienced a lot of foreign input in the historical period. Ireland did not begin to gain its independence from England until the 1920s.

Secondly your claim regarding Ireland that "its U106 population is self-evidently mostly the product of outside, non-Irish input and settlement." is just that an assertion only, it is clearly not demonstrable in terms of empirical evidence, suggested by the use of the term 'evidently' as there is no evidence to determine one way or another. As noted previously we have ultimately only speculation (albeit informed) the rest, including scales of probability, is built upon that foundation. In that factual context we can equally speculate and consider that along with some R U106 arriving via incursions and settlement from England, the Haplogroup had every opportunity to enter Ireland as part of an earlier admixture.

It is an assertion based on the evidence, which is considerable. Your notion, that U106 entered Ireland very early, is not its equal, having nothing to support it accept your assertion that such a thing was "possible". I agree. It was possible. It just wasn't likely and so probably did not happen.

Regarding the observation that: "Reservations or not, the inclusion of non-Irish surnames in the calculation of "Irish" U106 variance renders that calculation invalid for the purpose to which you are putting it, i.e., as some kind of indication of the age of U106 in Ireland. An invalid variance calculation cannot then be transferred over to England."

Such a selective interpretation is semantic sophistry in that  your original criticism to dismiss Tim Janzen's findings was that he used data, which you assert (again you do not know for certain) was drawn from a majority of supposedly English names. That was the contention you presented to claim his results were invalid in terms of R U106 variance in Ireland and the implied conclusions of it arriving at an early stage. If (and that is another significant  area of doubt in the critique) such names were undeniably English then the only aspect of the data that is changed or in any way less  exact is that fact it would then apply to people whose ancestry (according to your evidence free claims) was from Britain. That applying, then the questions concerning R U106 variance would indeed simply be shifted to Britain with the same question and considerations relating to its possible pre 'Germanic' arrival there.

It wasn't "semantic sophistry" (it would help if you would use terms you actually understand). I said exactly what I meant.

An "Irish" variance calculation that includes non-Irish surnames cannot be valid for the purpose for which you are using it, i.e., to establish the possibility that U106 entered Ireland in ancient times.

Such a calculation could not simply be shifted to Britain. The few Irish surnames it includes would have to be excluded first and the other British haplotypes would have to be included.

Another thing you just don't seem to understand is that variance does not establish a date for the arrival of a y haplogroup in a place. It merely sets an upper bound for the possible age of the haplogroup there. If variance says it could be 4k years old in a place, well, it cannot be any older than that there (barring something extraordinary like a bottleneck), but it could certainly be a lot younger.

Mike's calculation of North American U106 variance is a classic illustration. It is "high" compared to most places in Europe, yet no one believes U106 arrived in North America in ancient times.

Variance cannot be the sole consideration. History, the distribution of the haplogroup, and its known ethnolinguistic affiliations have to be taken into account.

Your entire argument posits no evidence of an actual movement of a U106 population to Ireland in ancient times. Instead, your entire position has been that Ireland's U106 variance is old enough that some U106 could have gotten there early. You seem to conclude that because that could have happened, it probably did happen. And that even though your vaunted "high variance" is based on a flawed calculation that includes the haplotypes of people who are pretty obviously non-Irish.

But there is no evidence that U106 entered Ireland early. None. U106 is relatively rare in Ireland and reaches its highest frequencies in the places where foreigners settled. We know from history that Ireland was settled and even ruled in the historical period by peoples from places with a lot more U106 than Ireland itself has.

All of those things are evidence that militates strongly against an early U106 arrival in Ireland. No bald assertions. No "semantic sophistry".

[rms2]

. . .

The main problem I see with looking at Irish surnames is that quite a large number of Irish surnames were anglicized over the centuries. Some surnames in Ireland are undoubtedly English or Scottish in origin, and some are undoubtedly Irish in origin, but I believe there are quite a few which could be either. I have a very Anglo-Saxon surname, but I know there is an Irish surname which has been anglicized into mine. People with that surname in Ireland could have either an Irish or an English origin. It isn't possible to make that determination from the surname alone.

That just means it isn't always easy to sort some English and Scottish-looking surnames that may possibly (but not certainly) be of Gaelic Irish origin.

But if the effort to exclude non-Irish surnames from a supposedly "Irish" variance calculation is not made, then that variance calculation is not reliable for the purposes for which it is being used here, i.e., to lend credence to the idea that U106 could have arrived in Ireland very early.

That is no problem for me, actually. I agree, it could have arrived very early, even in ancient times. I just see absolutely no reason to believe it did and lots of reasons to believe otherwise.

[whoknows]

Again I shall ignore the sarcasm,appeal to ridicule and other fallacious posturings, as the subject in hand is not about 'personality' but a reasonable speculation that R U106 may well have entered Ireland at an early stage, before Germanic culture had emerged.

Discussions for or against that scenario are equally dependent upon an absence of actual hard evidence, and rely instead on informed interpretation drawn from statistics relating to currently known data, that may or may not reflect accurately an ancient condition. In that regard I feel your comments are an articulate representation, when you assert:

"That is no problem for me, actually. I agree, it could have arrived very early, even in ancient times. I just see absolutely no reason to believe it did and lots of reasons to believe otherwise." (emphasis added)

On closing may I again request that this exchange be concluded on a mature, non hostile and balanced tone, personal attacks, veiled insults are entirely unhelpful. I genuinely value your freedom to hold a different belief on this issue and trust your can respect my right to take an alternative perspective.

[rms2]

Again I shall ignore the sarcasm,appeal to ridicule and other fallacious posturings, as the subject in hand is not about 'personality' but a reasonable speculation that R U106 may well have entered Ireland at an early stage, before Germanic culture had emerged.

You are right about that. I apologize and will try to refrain from being sarcastic. You should do the same, however, and quit talking about things like "semantic sophistry" and "fallacious posturings", which are both untrue and essentially meaningless.

Discussions for or against that scenario are equally dependent upon an absence of actual hard evidence, and rely instead on informed interpretation drawn from statistics relating to currently known data, that may or may not reflect accurately an ancient condition. In that regard I feel your comments are an articulate representation, when you assert:

"That is no problem for me, actually. I agree, it could have arrived very early, even in ancient times. I just see absolutely no reason to believe it did and lots of reasons to believe otherwise." (emphasis added)

The tremendous preponderance of the evidence is against an early entry of U106 into Ireland. We've been over it again and again and again, so I won't list it in this post (I may repeat it again in a subsequent post, however).

I'm not sure what you mean by "hard evidence", unless by that you mean some actual ancient y-dna. No, we don't have any of that yet, and we may have to wait awhile for it. I don't know of any ancient y-dna testing that has gone as far up the tree as U106. We're lucky if we get a short haplotype that can be predicted to be R1b.

On closing may I again request that this exchange be concluded on a mature, non hostile and balanced tone, personal attacks, veiled insults are entirely unhelpful. I genuinely value your freedom to hold a different belief on this issue and trust your can respect my right to take an alternative perspective.

Again, I will try to sound less hostile, but you need to take your own good advice, as well. Your posts have not been entirely free of the types of things you indict above.

Let me ask you a couple of questions.

Honestly, do you really believe U106 entered Ireland in ancient times?

If you do, why?

Note that these questions have nothing to do with "could have". I think we all agree U106 "could have" entered Ireland in ancient times. It could have entered China in ancient times, as well. But did U106 actually enter Ireland in ancient times and, if so, what are the reasons for believing that it did?