McCarthy Surname Study - Results

Last updated 28 March 2014


For a study of McCarthy genetic genealogy in relation to other Munster surnames please go to  

For further information and analysis of the Irish Type II haplotype of McCarthy Group A please go to

For postulated phylogentic trees for McCarthy subgroups too large to be presented on this webpage please go to



Two major clusters, one with a common ancestor who lived near the end of the Iron Age (McCarthy “Group A”) and the other several centuries later than this (“Group B”), account for about 50% of all McCarthys.  It is often the case that the progeny of chieftains flourish at the expense of the more impoverished 'serfs and soldiers' who surround them, although the high mortality rate arising from both inter-clan and internecine feuding involving Irish chieftains is an argument against this. However, if the progeny of Cárthach, King of the Eóghanacht Cashel (d 1045), did flourish, it seems likely that it will be found in McCarthy Group A or Group B.  According to the ancient genealogies, the first independently named branch of Cárthach's progeny gave rise to the McAuliffe surname, and those few McAuliffes who have tested their Y-DNA (outside of this Study) clearly have the Irish Type II haplotype associated with McCarthy Group A.

Whereas it is argued below that all the Group B McCarthys (who account for approx. 28% of all McCarthys) could have a common ancestor born in the era of Cárthach, it currently appears less likely that such a claim could be made for the approx. 22% who make up the Group A McCarthys, there being various lineages within Group A which probably only have common ancestry some centuries earlier. Moreover, claimants to MacCarthy Reagh ancestry are also found in McCarthy Group B (notably the McCarthy Rabagh subsept) and Group M (where there is also found a claimant to MacCarthy Glas, or Sliochd Feidhlimidh, ancestry). Although there are the occasional O'Sullivan and O'Mahony, and notably, (O')Calla(g)han who share the late first millenium ancestry of McCarthy Group B, marked with the occurrence of SNP R-L362, there is none of other prominent surnames deriving from the Eóghanacht Cashel with this ancestral haplotype, which McCarthys dominate. 

On the current evidence, this Study concludes that Group B McCarthys are the most likely to be direct descendants of Cárthach, and thus that either the Eóghanacht kings of Cashel, and sometimes of Munster, tracing back to Conall Corc, the reputed first King of Cashel, were of a lineage associated with SNPs P314.2 and L362, or that there was a break in the continuous line claimed by the ancient genealogical tracts. The Group A McCarthys share common first millenium ancestry with many whose surnames purport to derive from the Eóghanacht of Cashel (besides some whose names do not immediately suggest an Irish origin at all), thus even if they are not direct descendants of the eponymous Cárthach, their earlier ancestry is consistent with the genealogical concepts of the ancient tracts.

Groups C and E each comprise about 7 % of McCarthys. Group C shares common ancestry with the “Irish Type I” peoples many of whom have roots in the north of Ireland, Scotland and the north of England. Whether the progenitors of Irish Type II (Group A) and Irish Type I (Group C) ancestral haplotypes were born in Ireland or lived elsewhere in Western Europe prior to migration of their progeny to Ireland is open to debate and ongoing study.

Group D shares common first millennium ancestry with the main Kerry group of O’Sheas, considered to be descendants of the Corca Dhuibhne peoples who had occupied the shores and hinterland of Dingle Bay.

Group E comprises nine participants who are predicted to test positive for R-Z255. One has tested positive for this but negative for SNP R-L159.2, but it is possible that most of the others, who have more in common with one another, could test positive for L159.2.  The STR pattern associated with this profile is known as the Irish Sea (or Leinster) modal haplotype, and is also seen in Norway. The first certain McCarthy Crimen (Crimeen) to test is within this group.

One kit in Group F has tested positive for SNPR-Z253 and the others, closely related, are predicted likewise; the Sowney agnomen features both in this group and independent group D.

The remainder of McCarthys belong to numerous minor clusters or lineages unrelated to Groups A, B, C, D, E or F later than the Bronze Age (although further testing of Group M4 may prove it should be included in Group E) and are seen as likely making up the entourage of the tribe(s) which took the Mac Carthaigh surname from Cárthach, or were adopted or otherwise came to belong to the McCarthy family subsequently as suggested in 'McCarthys in Antiquity' on the 'Background' page of this website.  It must also be considered that there was more than the one Cárthach whose progeny took the Mac Carthaigh surname, just as is clearly the case for the O'Donoghues and several other surnames prominent in Munster.    


The results of y-DNA testing for haplotype and haplogroup determination are shown on the "y-DNA Results" page on this website, and indicate that a small number of McCarthys have been tested or predicted as being within haplogroups E1b1b1, G, I1, I2a, I2b1 and R1a1. However, FTDNA predicts that about 90% of McCarthys have common ancestry in the progenitor in which SNP R-M269 occurred (current FTDNA subclade designation R1b1a2). A number of geneticists believe he lived in the region of the Caucasus about 8,000 years ago (+/- a few thousand years!).  It is likely that most if not all of these McCarthys will also share common ancestry in the more recent progenitor of the subclade marked by SNP R-L21 (currently designated by FTDNA R1b1a2a1a1b4). As explained under “McCarthys in Antiquity” on this website, the haplotype estimated as occurring in this common ancestor of over 70% of Irish men is known as Atlantic Modal Haplotype (AMH). (This is sometimes prefixed “Super Western”, thus SWAMH). In the discussions of McCarthy Groups A-F below, sequences of DYS mutations defining the “modal haplotypes” of specific lineages in the progeny of the R-L21 ancestor are referenced to AMH. R-L21 is estimated to have occurred about 3,700 to 4,200 ybp (years before present) in a foetus created somewhere in mainland Western Europe. It appears that two other SNPs, R-L459 and R-Z245 occurred in conjunction with R-L21, or in a generation or two before or after in the same ancestral lineage, and that SNP R-DF13, identified in Spring 2012, in the common ancestor of about 95% of those carrying R-L21, the implication being that DF13 probably occurred just a few hundred years later than L21.  Initial testing indicates that all McCarthys who test positive for R-L21 will also be positive for R-DF13.

The 2004 analysis of McCarthy Surname Study results, provided below for reference, had only the benefit of the 12 FTDNA Panel 1 markers, but was able to divide the R haplogroup McCarthys into three groups, the third being a miscellany. Groups A and B labels have been retained to this day. The current R-M269 groups (A-F and M) are more precisely defined below. 

The diversity among present day McCarthys is illustrated by the following breakdown of haplotypes for the 66 McCarthy participants in Prof. D G Bradley and B McEvoy’s Trinity College, Dublin study of 2003/04 and results from the McCarthy Surname Study. The latter includes data taken from sources other than FTDNA; where participants share very recent common ancestry, they are counted as a single entry. A comparison with data inclusive of others of Irish origin from the Ireland yDNA project is also provided.

HAPLOGROUPS E1b1b1, G, I1 and I2b1

For information on the origins of the haplogroups other than subclades of R-M269 indicated in the above table, you are referred to the excellent accounts on the Eupedia website, at:

Similar percentages to those found in the FTDNA McCarthy Study for E1b1b1, G, I1, and I2b1 are seen in the overall statistics for Ireland.  E1b1b1 is an eastern Mediterranean marker.  I1 proportions in Scandinavian countries account for 30 - 40 % of the male population, of a similar magnitude to R1b, and thus it is quite possible that this entered Ireland with the Vikings.


For information on this haplogroup please refer to the I2a Project website at

The two McCarthy participants in this group appear to belong to Isles Groups A1 and C2 respectively, although that of the latter has only tested 12 markers. Pending discovery of more recent SNPs, I-L161 is predicted as the terminal SNP for both, but occurred many thousand years ago.

Isles Group A1 appears to be that of almost all, if not all, the I2a O’Driscolls, and accounts for 55% of this surname. All but one of the O’Driscolls who have tested to 111 markers distinguish themselves from other Isles Group A1 participants with an allele count of 9 at the slow mutating Y-GGAAT-1B07 (rather than 11 seen elsewhere in this group). It will be interesting to see if other surnames of West Cork origin, and particularly McCarthys, in haplogroup I2a replicate this.

The O’Driscolls claim descent from the Corca Laidhe tribe of southwest Cork. The McCarthy Isles Group A participant currently has family lore of belonging to the Clan Teige Ilen McCarthy sept, who occupied lands immediately to the north of Skibbereen,  Historical works give no reliable origin for this sept. The McCarthy participant has no matches nearer to any O’Driscoll than 32 / 37 and this suggests his most probable point of common ancestry with the O'Driscolls could be as early as the time surnames were coming into use. 

Agnomens positively identified as pertinent to McCarthys in Haplogroup I2a:    Clan Teige Ilen


The Irish Type II haplotype is characterised by the following departures from SWAMH: DYS 391 = 10, DYS 385b = 15 or 16, DYS 439=11, DYS 447 = 24, DYS 456 = 15, DYS 442 = 13 and, where tested, DYS 565 =11. Where testing has been extended to 111 markers, the following set of mutations from SWAMH also characterise this lineage: from 12 to 11 at DYS 636, probably 20 to 21 at DYS 712 (a fast mutating marker, so not a reliable indicator), 13 to 14 at DYS 532, 17 to 16 at DYS 504, 23 to 24 at DYS 635, 17 to 18 at DYS 510. These DYS STR values are consistent with the “modal haplotype” known as (South) Irish Type II. 

In many sub-branches of Irish Type II, DYS 456 = 16 is found alongside the above mutation from SWAMH. DYS 456 appears to have a high mutation rate when oscillating between 15 and 16, but is less susceptible to further mutation to 14 or 17. It does appear that a mutation from 16 to 15 occurred soon after the time of SNP R-L21 in the Irish Type II lineage, but because of the scattered occurrence of both 15 and 16 the value at DYS 456 is, therefore, not a useful indicator of formation of any particular ancient sub-branch of the Irish Type II tree. 

Similarly, while an apparently early mutation from 12 to13 at DYS442 is a valid identifier (in conjunction with other mutations described above) of the Irish Type II haplotype, the value 12 is also common among most surnames,with no evidence that this can be explained by a single back (13 to 12) mutation event common to much Irish Type II ancestry. While further mutation to 11 is, in Irish Type II haplotypes, rare, it may be that this DYS is particularly prone to back mutation from 13 to 12.

DYS 565 and DYS 636 have low mutation rates. Subsequent mutations at these DYSs are therefore far less likely, and their allele counts of 11 are, in tandem, a reliable indicator for Irish Type II where an SNP test is positive for R-L21 or R-DF13.

For information on Irish Type II SNPs please go to the Subclades of R-L21 section on the Project Background page.

Over 50% of O’Sullivans and southern O’Donoghues, 50% of O'Keeffes and over 30% of O’Mahonys, all names claimed to be of Eóghanacht origin, have (South) Irish Type II haplotypes. The reduced percentage (a little over 20%) for McCarthys indicated in the tabulation above is largely due to the preponderance of the P314.2 lineage (see Group B below), which is not seen in significant numbers in other names associated with Eóghanacht origins, although the O’Sullivans are represented in two separate branches of the P314.2 lineage.

However, Irish Type II haplotype is also observed in both surnames derived from Munster tribal origins other than the Eóghanachta and in many surnames associated with other parts of the ‘Isles’. This might be explained in some cases by an efflux of Munster peoples through marriage alliances, trading and warring who did not return, and therefore carried their haplotypes into other parts of Western Europe in general and the Isles in particular, often places where surnames did not come into use until much later. But an alternative explanation is that the progenitor of the modal Irish Type II haplotype was born not in Munster but in England or elsewhere near the Atlantic seaboard, his progeny subsequently dispersing principally within the Isles, and flourishing particularly in Munster. Under such circumstances Irish Type II could have arrived in different parts of southern Ireland in separate migrations each resulting in fresh tribal configurations.

Mike Walsh, in his excellent work on the analysis of the numerous identifiable lineages irrespective of surname since the occurrence of SNP R-L21 (ref., has identified a subgroup of Irish Type II (labelled 4466-T2-C and accounting for 14% of all Irish Type II) which shares the following mutations: 24 to 23 or 22 at DYS 390, 15 to 14 at DYS 607, 22 to 18 at DYS 481 and 18 to 19 at DYS 510.  There are few, if any, Munster surnames in this subgroup, although a McCarter in our project belongs to it, and the spread of surnames and their ancestral locations indicates that the subgroup flourished in Ulster or Scotland. The ancestors of this subgroup could of course have migrated from Munster, but it could also be that they never set foot in southern Ireland. 

Estimates of the date of birth of the progenitor of the fully-fledged (ancestral) Irish Type II haplotype appear to focus on the early to middle part of the first millennium A.D., so it does seem feasible for the Irish Type II O’Sullivans, O’Donoghues, O'Keeffes, O’Mahonys and McCarthys to all have common ancestry in a powerful figure in the Eóghanacht hierarchy. Several separate subgroups (or clusters, or lineages) of McCarthys are detectable within McCarthy Group A, each with their own progenitor who probably lived several centuries before the introduction of the McCarthy surname (in the 11th century A.D.). However only one of these lineages, if any at all, can claim descent from the Cárthach, King of Cashel, to whom the ancient genealogical tracts attribute the derivation of the M(a)cCarthy surname.

postulated phylogenetic tree, showing all Irish Type II McCarthys who have tested to 37 STR markers or beyond, as well as principal clusters for the Irish Type II portions of other Munster surnames which have been tested in sufficient numbers, is provided at . Please read the Important Notice thereon, and note that no attempt has been made to integrate the trees for each surname beyond the establishment of the ancestral Irish Type II haplotype.

gnomens / branches identified as pertinent to McCarthy Group A:  Brogach, Cruig, Guideach (Gaibhdeach), McAuliffe MacCarthys, Muineagh


McCarthy Group B shows most or all of the following departures from SWAMH: DYS 388 = 13, DYS 390 < 24, DYS 389-2 less DYS 389-1<16, DYS 439 = 13, DYS 464 = 15-15-16-17, DYS 460 = 10, and, where tested, DYF 395S1a =16, DYS 406S1 = 11, DYS 557 > 16 and DYS 617=13. From the few results where 111 DYS markers have been tested, the following set of mutations from SWAMH in FTDNA Panel 5 (its 68th– 111th markers) also characterise this lineage: 18 at DYS 650, 13 at DYS 513, 25 at DYS 552. This group tests positive for SNP R-P314.2 (and thus for its predecessor R-DF21, discovered in 2011), and, so far, for R-L362, although more testing by McCarthys is needed to verify this.

This is the largest grouping of McCarthys, accounting for almost 30% of all of us. Fortunately, four of the mutations from SWAMH which contribute to the modal haplotype fall within FTDNA’s Panel 1 set of DYS locations, so that the cluster was identifiable by the original Administrators of this McCarthy Surname Study in 2004. However the added bonus of the lineage having its own specific SNP, R-P314.2 was only discovered through a “R-L21 Walk through the Y” (WTY) test late in 2009. Then late in 2010 a further, younger, SNP in this lineage, R-L362, was also incidentally discovered in the FTDNA laboratories in two samples, one of which belonged to a McCarthy member. 

At 17 February 2013 the McCarthy Study was aware of 33 positive tests, among all surnames, for SNP P314.2 and a further 115 members of various projects whose distinct haplotypes indicate they too are almost certainly positive for this SNP.  McCarthys account for 33 out of the total 148 (22%). Seven people are also known to have tested positive for SNP L362: four are McCarthys but the proportion of McCarthys in this subclade is predicted to be considerably higher than 50%.  A McCarthy from each mini-cluster on the “P314.2 tree” should test for SNP L362.

The other surnames involved in the P314.2 lineage are predominantly Irish (though with low representation from Leinster), and Scottish. The frequent occurrence among McCarthys is not reflected in other Munster surnames: O’Sullivans and O’Mahonys each have about 3%  P314.2 lineage and there are no O’Donoghues at all in the FTDNA O’Donoghue Surname Study, although there were one from Munster and one from Leinster in the 2003/2004 Trinity College, Dublin Project. (Note: the Trinity College study indicated a much less diverse spread of names outside Munster with the characteristic haplotype, with six Heaneys being the only occurrence reported with Ulster or Connacht origins and nine occurrences of Kennedy, a name normally associated with the Dalcassian counties of Clare and Limerick and their bordering territories).

There is also a sprinkling of English names among the 148 tested or predicted positive for P314.2, then one from Norway, one from Normandy and one Sicilian suspect. These last three are the only ones without the characteristic 13 at DYS 388, a DYS which mutates very rarely, but the first two have tested positive for P314.2. Their haplotypes are more diverse, pointing to a more distant common ancestry with the more closely related Irish and Scots representatives. For those with DYS 388 >12, all but a very few appear to fall into two parallel subgroups, one identifiable with a 12 to 13 mutation at DYS 617 followed usually by 16 to 17 at DYS 557, and the other with a 19 to 18 mutation at DYS 448 and possibly, where tested, 25 to 24 at DYS 714 (in FTDNA Panel 5).

A postulated phylogenetic tree, showing the inter-relation of tested or predicted P314.2 kits of all surnames, is provided at . This is the basis of the analysis which follows below. This includes all McCarthys who have tested to 37 or more markers. It will by no means be precisely correct (although the greater the extent of testing to 111 markers the more reliable it becomes), but it should be indicative of the development of the "P314.2 lineage" as part of the R-L21/DF13/DF21 family.  The tree classifies major historical branches as P314.2 “Type” numbers. Inclusive of postulated back mutations, it indicates an average of 21 STR mutations between the occurrence of R-L21 and the present day for those McCarthys (and a similar number for other surnames) who have tested to 67 markers. There appear to be no STR mutations between the occurrence of SNPs L21 and DF21, indicating that both DF13 and DF21 occurred within a short time of L21.  Estimates by others suggest DF21 occurred about 3,600 ybp (about 1,600 B.C., early in the Irish Bronze Age). Thus this implies, on average, one FTDNA Panel 1-4 STR mutation every 171 years, which is broadly consistent with more scientific analysis of Y-DNA data by others, and this is the basis of the timeline below. However, it is essential to appreciate that

·        this age for DF21 may be subject to debate (although a few hundred years either way makes little significant difference to the timeline and of course does not affect the sequence).

·        STR mutations occur randomly, not at pre-set intervals! This is immediately demonstrated by the fact that although McCarthys have an average 5 FTDNA Panel 1-4 STR mutations following the postulated 26 to 25 back mutation at DYS 447 (i.e in the past 860 years), the range is from 1 to 9.

·        the simplistic but transparent timeline estimates are intended for guidance in the postulated sequence of events and possible correlations with genealogical data in the ancient tracts and documented movement of peoples. While there is confidence in the sequence of events, their actual dating may be adrift by a century or two.

The P314.2 types are as follows:

·        Type 1        DYS 388 = 12

·        Type 2        DYS 388>12 and DYS 617 =13

·        Type 2A      As type 2 and L362 positive

·        Type 2B      As type 2 and L362 negative

·        Type 3        DYS 388>12, DYS 617 = 12 and DYS 448 = 19

·        Type 4        DYS 388>12 and DYS 448<19 (subject to consideration of feasible back mutation to 19 at the latter).

·        Type 4A      As type 4 and with DYS 458 = 17 and DYS 576>19

·        Type 4B      As type 4 but with DYS 458<17

The timeline derived on the above basis is as follows:

      1600 B.C.          Assumed date for occurrence of SNP R-DF21.

      1600 - 230 B.C. Period in which the first 8 FTDNA Panel 1-4 (first 67 markers) STR mutations and SNP 314.2 occur.

      30 A.D.             Progenitor of P314.2 Type 1 born (Norwegian and Benelux kits).

      370 A.D            Progenitor with 12 to 13 mutation at DYS 388 born.

      540 A.D            Progenitors of Type 2 (with 12 to 13 at DYS 617) and Type 4 (with 19 to 18 at DYS 448) born.  Progenitors of Type 3 (any without either of these mutations after consideration of possible later back mutations) born at or after this time.

      630 - 970 A.D    SNP R-L362 occurs in Type 2A progenitor.

      710 A.D            Progenitor of Type 4A (Martin / Keenan branch) and common ancestor of all Type 4B subgroups born.

      890 A.D            Progenitors of the various subgroups under Type 4B (Munster branch 2, possible Munster branch 3, Scottish branches, etc.) born.

     1060 A.D           Progenitors of the four branches under Munster branch 1 born

     1230 A.D           Progenitors of the numerous Munster branch 1 subgroups having common ancestry in the postulated 26 to 25 back mutation at DYS 447 born at or after this time.

So how is the distribution of the “SNP P314.2 lineage” to be explained? The first question is: where did its carriers live prior to the occurrence of the DYS 388 mutation? Had they been in Ireland for centuries, the Type 1 carriers found in continental Europe being explained by early Iron Age seafaring ventures and this more diverse type having subsequently died out in Ireland itself? Or was P314.2 born in continental Europe and brought to Ireland shortly after or at some time before the occurrence of the 12 to 13 mutation at DYS 388? Until and unless we locate more Type 1 carriers we cannot answer this question, and of course we may yet find Type 1 in Ireland. However, if the latter, the timing seems valid for arrival in such tribal groupings as the Erainn, the Brigantes, the Belgae and the Fir Bolg (if these last two were not one and the same) and other tribes opting to seek new pastures as a consequence of Roman expansion in Western Europe and, from 43 A.D, into England. For example, finding early, diverse, forms of the haplotypes associated with P314.2 in some distant land such as Greece (when it could support one Fir Bolg myth), or in others in Northern France and the Benelux countries, and later in south-east England, whose ancestors had not migrated in the subsequent 2,000 years (supporting the Fir Bolg = Belgae suggestion) would give some clarification.

In an earlier analysis on this website it was noted that the three Type 1 kits were all from locations associated with the Vikings, implying that P314.2 could have had Norse origins. Normandy, like Ireland, Scotland and parts of England and Wales, was subject to significant Viking invasions, and Sicily (and the south of Italy) was subsequently overrun by the Normans. Whatever the margin of error in the above timeline, it clearly shows that this theory is inconsistent with the genetic evidence now available, although a much earlier Norse origin is still possible.

Many lineages of 3,000 years ago will simply have died out for whatever reason, and others hung by a thread. P314.2 Type 3 apart - the tree here shows just two strands each with a single participant, and postulation of appropriate back mutations could place either in a Type 4 branch - the two mid first millennium progenitors in which the DYS 448 and DYS 617 mutations occurred appear to have been the sole ambassadors of surviving lineages of P314.2 in Ireland and / or Scotland somewhere around the middle of the first millennium A.D.  One explanation for such a pinchpoint in the survival of any lineages in this era is offered by the climatic disaster of the 530s and the plague which was endemic in Ireland and beyond for decades thereafter, substantially reducing populations, so that at the dawn of the seventh century A.D. many earlier lineages had failed to survive, but expansion of those that had was beginning to get underway.

Did P314.2 spread into Munster from elsewhere in Ireland or Scotland, or vice versa?  If the former, Type 2 must have arrived first, probably before the occurrence of the DYS 617 mutation but certainly very soon after: note that just a Higgins (Type 2B) has this mutation, is certainly negative for SNP L362 and has a midland / northern Irish name, and there are no Scottish names here. (There is paper trail evidence which supports the possibility of much later NPEs accounting for some of the English names).  An argument is presented below for Type 3 being located in Munster by the middle of the first millennium. Such a migration southwards could have brought Type 3 with it, or it could have followed soon after (in either case the DYS 448 mutation would have occurred in the north of Ireland or Scotland). Those Munster branches of Type 4B would then have arrived in Munster perhaps 350 years later, but in time to be assimilated into tribes which were soon to take surnames that we know today.  Around 100 B.C., some branches of the Erainn in the north of Ireland are said to have migrated to Munster. Some may claim this as an explanation of the distribution of P314.2 surnames, but while some potential error in the accuracy of the above timeline is acknowledged, such a migration is considered far too early to be relevant to this analysis.

While a migration from Scotland or the north to the south-west of Ireland certainly remains a possibility, the arguments for dispersion from a Munster base are more convincing. The DYS 617 mutation, which it has just been argued probably occurred in Munster, appears to have been next in line following that at DYS 388, thus allowing only a small window for migration into Munster if that were to be the case: i.e it is more likely that the progenitor of the DYS 388 mutation was already in Munster.

The parallel DYS 448 mutation has given rise to both Munster and non-Munster progeny, but if Munster was its home, it is suggested only the future Martins and Keenans of Type 4A could have moved out of Munster in the ensuing few centuries. Since Type 4B contains both Munster and non-Munster branches, it is suggested that any migration of the latter, such as those who came to form or at least be assimilated into Scottish clans, did not occur before the 17 to 16 mutation at DYS 458, somewhere in the second half of the first millennium. This was a time of Eóghanacht expansion in Munster and certainly links between Munster and distant Scotland were well established. The ancient tracts record that Conall Corc, the 5th century first Eóghanacht king of Cashel (in present day Co. Tipperary), took as his second wife the daughter of the king of the Picts in Scotland, and as a consequence at least two of his sons are said to have established Eóghanacht septs in Scotland. Although possibly too early to account for our Type 4B migrants, this might be seen as the forerunner of later migrations. Seen in this light, the phylogenetic tree certainly supports a theory proposed many months ago by Keith Martin (in whose WTY test SNP P314.2 was discovered), viz, arrival of P314.2 in Munster about the time the Erainn tribes are first recorded and later dispersion through maritime trading, or, as highlighted here, marriage alliances.

One other tantalising possibility mooted by Keith Martin should be mentioned here. By the middle of the first millennium, the tribal grouping of Mairtine are recorded as occupying a region including present day Co Tipperary. This same region was fast becoming the stronghold of the emerging Eóghanacht tribes who came to dominate Munster and from whom the ancient genealogical tracts indicate the McCarthys, (O’)Sullivans, Dennehys and Dugans (among others) descended. Were the Mairtine assimilated into the Eóghanachta? Are the Martins of P314.2 Type 4A descendants of the Mairtine, in which case this would:

(a)  indicate Munster, and more specifically Co Tipperary, as the base for the spread of P314.2 in the second half of the first millenium.

(b)    explain how the P314.2 lineage became one of the haplotypes of the Eóghanachta.

(c)    potentially place those with SNP R-L362 in Cashel, Co Tipperary, the seat of many Eóghanacht kings in the second half of the first millennium, including the Cárthach whose name has given rise to this Study.

However there is another twist with this theory. While the Mairtine disappear from history about this time, while some sources link them to the Corca Laidhe, the predecessors to the Eóghanachta in the dominance of West Munster, others suggest they might have earlier arrived in the Co. Tipperary region from Connacht or Ulster, supporting the theory of the initial spread of DYS 388 = 13 from the north into Munster, but then followed by a dispersion from Munster as now being proposed. In either case it has to be acknowledged that there appears to be a gap of some 400 years or more between the disappearance of the Mairtine from the written records and the appearance of Martin as a surname, and this is not supportive of these theories.

Since Type 2A (with SNP L362, subject to more proof from testing) has come to be the dominating P314.2 lineage in Munster, and since it is so often the progeny of chieftains who flourish, it is not unreasonable to propose, if the foregoing arguments are valid (even without the Mairtine consideration), that the progenitor of L362 could have been a powerful force in the Eóghanacht hierarchy, and of Cárthach himself.  The ancient genealogical tracts indicate that many McCarthys, O’Sullivans, O’Mahonys, O’Donoghues and O’Keefes share a common first millennium A.D ancestry and genetic evidence suggest this is the Irish Type II ancestral haplotype of McCarthy Group A. However, the Group A McCarthys comprise a number of strands whose common ancestry cannot be as recent as the time of Cárthach (d 1045) whereas the common ancestry of all the Group B McCarthys could be as recent as this.  Two other examples of circumstancial evidence supporting such a possibility follow.

(1)    Ancient genealogies claim that the O’Sullivan and Dennehy surnames shared common ancestry in Finghin, a King of Munster in the line of the Eóghanacht of Cashel who died about 618. This is broadly consistent with the detail shown in the phylogenetic tree for Munster branch (II), and although the ancestry of present day O’Sullivans and Dennehys certainly shows other ancestral haplotypes to be dominant, it remains possible that Finghin and thus his brother Failbhe Flann, a purported direct ancestor of Cárthach, were of the P314.2 lineage.

(2)    The (O’)Calla(g)han surname is said to derive from Cárthach’s great grandfather, Ceallachán of Cashel, King of Munster, who died about 954, or Carthách’s cousin of that name.  The Callahan DNA project has just one possible Irish Type II participant, but a substantial number of Irish Type I (McCarthy Group C) and of Irish Type III (of which the McCarthy Study has none). It also has at least four P314.2 lineage haplotypes, and if Cárthach truly belonged to McCarthy Group B, reference to the postulated P314.2 tree shows that these are perfectly placed to be descendants of Cárthach’s cousin Cellachán.

With regards the McCarthy agnomens tested to date, McCarthy Group B has two of McCarthy Rabagh, said to be a subsept of MacCarthy Reagh, which in turn traces back to Cárthach according to the ancient genealogies. It also has a McCarthy Tallin, whose family lore is of descent from the Reagh dynasty and a McCarthy Norsa, indicating descent from Teige an Fhorsa, of the Sliochd Feidhlimidh. This latter sept is also recorded as being a branch of the MacCarthy Reagh family; however it is synonymous with the MacCarthy Glas sept, and indeed Daniel MaCarthy Glas published a book on the sept and its origins in 1880. However the one representative with the Glas agnomen has a completely different haplotype, with common ancestry with another McCarthy estimated as being prior to 1800. This suggests an NPE in the lineage, perhaps between 1600 and 1800.

The conclusion on the current evidence is therefore that P314.2 could well be an Eóghanacht, and prior to that Erainn, lineage, and that Group B McCarthys are the most likely to be direct descendants of Cárthach. If that is the case, whether the preceding Eóghanacht kings of Cashel carried SNP P314.2 in the unbroken line back to Conall Corc claimed in the ancient tracts, or there was at some stage an unrecorded break (or NPE) in this lineage, is open to speculation.

Agnomens identified as pertinent to Group B:  Farshing / Forshing (Fairsinn), Norsa (Sliochd Feidhlimid), Rabagh (Rabach), Tallin / Tollin.


McCarthy Group C is a small subset of the original Group C having most or all of the following departures from SWAMH: DYS 390 =25, DYS 385b = 13, DYS 392 = 14, DYS 464b = 16, DYS 607 = 16, and, where tested, DYS 413a = 21, DYS 534 = 17, DYS 481 = 25 and DYS 487 = 14. These values are consistent with the “modal haplotype” known as (North-west) Irish Type I, for which SNP M222 is an identifying marker. In 2011, SNP R-DF23 was found to sit between R-L21 and R-M222 on the phylogenetic tree, and subsequently a further SNP, R-DF49, has been shown to be upstream of DF23 but also downstream of L21.

Several recently identified SNPs may further subdivide the large M222 group. Ten of our members (7%) clearly share the (North-West) Irish [haplo]Type I, although only two have tested (and proved positive) for its associated SNP R-M222. This modal haplotype was identified as a consequence of the testing carried out under the aforementioned Trinity College, Dublin 2003/04 project and the common ancestor in whom the last of its characteristic mutations occurred was then estimated to have lived around the time of the 5th century Niall of the Nine Hostages, in the north-west of Ireland. It is shared by about 20% of the male population of present day Donegal and in significant percentages elsewhere in northern Ireland and southern Scotland. Its occurrence in McCarthys, and occasionally in other families of more recent Munster origin, may be attributable to a variety of causes, any of which could equally explain the presence of McCarthys in provinces other than Munster. Prior to more recent reasons for mobility, these include trading connections, hostage taking, alliances made between tribal groups, military forces not returning to their homelands and adopting local surnames  (such as remnants of Hugh O’Neill’s and Hugh Roe O’Donnell’s armies which had marched south to Kinsale in 1601), ecclesiastical obligations, or simply individuals seeking new pastures. If SNP P314.2 (see Group B above) had arrived in Munster by movement of peoples from the north / north-west of Ireland, certainly we would expect Irish Type I / SNP R-M222 to have arrived with it.

For more detailed information please refer to (and join, if you have this haplotype!) the R-M222 Project at and


McCarthy Group D has most or all of the following departures from SWAMH: DYS 385b >14, DYS 439 = 11, DYS 449 <29, DYS 464d = 18, DYS 576  <18, DYS 570 <17, DYF 406S1 = 11 or 12, DYS 534 >15, DYS 617 = 13, DYS 710 >35, DYS 650 <18, DYS 504 <17.

The five (possibly six) McCarthys in this group clearly share the same ancestry as the main Kerry group of O’Sheas, one of four principal O’Shea groupings who, much like the disparate McCarthy groups, only share ancestry with each other in the Bronze Age.  These O’Sheas test positive for SNP R-L513, and the one McCarthy who has tested for this has done likewise.  For a most informative history of these Corca Dhuibhne peoples go to

STR mutations which define the L513 haplogroup are the combination of 10 to 11 at DYF 406S1 and 12 to 13 at DYS 617. Thereafter three subclades marked by SNPs 705.2, L69.5 and P66, and L193 split off, each with its own set of differentiating STR mutations. The McCarthys of Group D, main Kerry group O’Sheas and others of predominantly southern Irish surnames follow a parallel path to these three subclades, but as yet no further SNP has been discovered to identify or subdivide them. A subsequent 11 to 12 mutation at DYF406S1 is seen in many O’Sheas and L513 McCarthys; this is postulated to have occurred long before surnames came into being as it is also seen in haplotypes of the same root for O’Gara, Kingston and others. Where DYF 406S1 = 11 (in one sizeable subgroup of O’Sheas defined by other common mutations and in one McCarthy) it is deduced this was due to a later back mutation to 11.  The following tree postulates the inter-relationship between these O’Sheas, the Group D McCarthys and two closely related Colemans, and major subgroups of the names O’Gara and Kingston.