Re: "..I submit that is a problem because the project admin should be intimately knowledgeable about the genealogy of their project surname and should be checking the lineages of their members. That is the reason I say projects for common surnames need to be broken up by haplogroup with the larger subgroups needing multiple co-admins, and that is why people shouldn't tackle more than one project, unless the surname is uncommon. IMO, we are not just taking tickets at the door and passively letting the members fend for themselves. We need to be deeply engaged in our projects." Speaking as one of those large-project administrators even with several co-admins, I submit the contrary. It appears that we have here the classic "blind men and the elephant" situation; each perceives a part of the animal and believes there is no more to it than what he feels. The first thing to be said about this elephant is that there is a qualitative difference between projects for common names (the "wall") and those for less-common names (the "rope"). A great difference lies in the presumption that people of the same name are related; for the several most-common surnames, it ain't necessarily so. To establish a relationship between two individuals with the same surname takes much more than them appearing in the same place at the same time. This qualitative difference makes traditional genealogy trickier, needing more evidence to meet accepted genealogical standards. Even then, conclusions must often be based on inferential, indirect evidence. (Preferably, with citations of the evidence and the logical processes for deriving conclusions.) Such standards-compliant genealogies are not only difficult to construct, they are hard and time-consuming to evaluate. (One could go on for pages about whether Zachary Taylor the US President was descended from Rowland Taylor the English Protestant martyr.) Often, evaluation will prove inconclusive. Surname origin makes a difference too. Is it "single-point", or "multi-point"? Did it arise in a narrowly-defined area or in many widely-spread places? Is it derived from a noble title or a common occupation? The former may have only a few separate lineages; the latter will have many unconnected lineages. For common, multi-point origin, surnames, one must speak in generalities about the origins, even after extensive research. There is no one story about the name, but many - mostly unknown - stories. (Disbelieve most of what you read on websites of "surname-history" providers; they are in the money-collecting business, not the history or genealogy business.) A third difference is in "open" vs. "admin approval required" projects. Open projects make joining easier, which has both good and bad aspects. Getting members is simpler; some of them will not find DNA matches within the project. (One then tries to ease the member into a more appropriate project.) As a project increases in size, there is a fourth qualitative difference; the role of administrator changes -- away from "boss" and toward "facilitator" or "mediator"; less hierarchical, more collegial. Instead of "owning" the project, the admin becomes the delegated representative of the members. He or she is entrusted to make certain decisions, but for the benefit of the members and project goals. "Deeply engaged" can have more than one meaning. A large-project admin for a common, occupational, multi-point origin surname may spend hours per day on issues that don't come up for other admins and approach the issues in a different manner. We may engage at a level some admins don't know exists. Genealogists -- as gen. society officers will attest -- tend to be a stubbornly opinionated lot, particularly for their own work. What almost no project member takes well is being told that his or her submitted genealogy is a piece of junk -- despite how junky it may be. (You may have noticed the rarity of objective evaluations of living authors' genealogies.) For an admin not to appreciate that human reaction almost guarantees the project will remain small. Small has consequences in terms of surveying surname DNA. Neither I nor my co-admins relish the role of "genealogy police", not one would accept the job. We do not censor nor vouch for trees submitted by members and make that clear to all. The accuracy of the tree is the submitting member's responsibility, though we make comments when the tree conflicts with DNA evidence. Should we break the project down by haplogroup? I submit that this is a confusing and unworkable proposition. An untested person does not know his haplogroup and, thus, which project to join. We may discourage members from joining any of the sub-divisions. We would wind up with an R1b project of almost the same size and other projects struggling for administrative support. This would benefit none of the parts. The Smith surname project has divided by geographic area -- Southern Smiths, Northeastern Smiths, Worldwide Smiths. This may be less confusing and more workable; at least, a prospective member can choose by earliest known ancestor. Possibly, division by country of origin would be better: Smith should not include Schmidt; Brown should exclude Brun & Braun; Taylor should not include Schneider or Szabo. (We don't have any.) OK, call it "compromise" if you will. I maintain that our project's approach is appropriate to the mission and the circumstances. It may not meet some expectations of how a project should run; but -- if the job is helping people find their ancestry via DNA -- it works. -ralpht_/)

Ralph, Thank you for a cogent description of the issues confronting a large surname project. Ours happens to be a patronymic. I agree with your blind man simile, although some will continue to insist that their tail is the animal, or in some cases, the hairs on the tail. . . We take a similar approach to what you have described. Our project has been through two lead administrators and I am the third, so I inherited a tradition of not requiring lineages as a condition of group membership. I suppose I could kick all the ones out that have not submitted lineages, but to what end and effect? I also have viewed my role as facilitator. We request lineages of all new members but have never required them for ordering a kit, and have fairly tight unpublished rules for subgroup membership, although there are exceptions. It would make more sense for us to require 37 or 67 marker tests than lineages, but that's not a fit with FTDNA's business model so not going to happen, either. We have taken the attitude that the DNA speaks for itself and that it is up to members to resolve contradictions. We have had more than one instance where the way the DNA chips fell, two or more individuals claimed descent from a common ancestor but clearly could not have been as they were in different haplogroups. In our most recent example, individuals A and B fit into existing subgroups with over five members each and no apparent link between them, contraindicating an NPE. We allowed the discrepancy to stand and the A member (the one with the weaker/less accepted paper trail), removed his suspected deeper lineage from his entry on the Results page. Two years later, a descendant of the immigrant who still lived adjacent to the family's original land grant was tested at the request of family B, and matched family A. This contradicted the lineages in two fairly well known books and other accepted sources. Researchers in family A have ideas about how the error occurred and both families are working on this. As for dividing the project by haplogroup, that still doesn't solve any problems in my view. Here are our (not exhaustive) stats: E1a: one singleton E1b1a: one group of two plus ten singletons G: one singleton I1: eight unrelated groups ranging in size from two to 14, plus five singletons I1d1: one group of ten I2a: one group of six plus five singletons I2b1: four non related groups ranging in size from two to 29, plus six singletons J1: one singleton J1a4b: one group of two J2: two singletons N1c1: one singleton Q1a2: one singleton R1a: two singletons R1a1: two groups with five and six members R1b1b2: 14 non related groups ranging in size from two to 17 members, plus 35 singletons This amounts to 15 haplogroups not counting the subgroups of R1b1b2, and at least 32 family groups. It would seem a little silly to have a surname project for a haplogroup singleton, of which there would be at least seven in the above list. A few of our family groups have active researchers that do work with new subgroup members to correct or extend their lineages, but they are not administrators. All our projects have slightly different circumstances and goals, and amounts of control the admins wish to exert. Large projects such as ours clearly require a different approach from one name studies or lineage projects. Don Jackson FTNDA Jackson Project Co-Administrator I1d1/J1b1a