Friday, December 28, 2012

Why I’m Creating a Junk Tree on Ancestry

No, it’s not revenge for all of the poorly documented trees I’ve had to wade through on Ancestry.

Nor is it a psychological experiment to see just what kind of ridiculous garbage people will copy into their own trees (grandmothers born after their grandchildren, etc.).

Nor is it a post-modern statement on the futility of associating ourselves with long-dead people.

It’s not even a decision to go over to The Dark Side of Sloppy Research because I’m just too darned busy and tired to bother with the Genealogical Proof Standard.

It has to do with DNA.

Specifically, the sets of DNA matches that I have been able to (tentatively) tie to specific lines.

Because some of my DNA matches appear to be in those “fuzzy” parts of my ancestral lines.

And some of the matches I am getting are interesting - no, make that very interesting.  And some of those tentative matches appear to be supported by other matches, some of which go even farther back along the same lines.    

So, here’s the deal.  I am not including any information that I know to be false.  I am, however, going back a generation or two farther than the documentary evidence with which I am familiar would cover, i.e., some of these are “reputed” parents (or parents of parents) that are commonly encountered in online trees or claimed in a family history, online post, etc., without citation of sufficiently thorough supporting evidence.   

I do not do this with my other, “main” Ancestry trees; additions are only made to these trees when I can document the connection.  However, for several reasons, these trees have not been useful to me with Ancestry DNA connections.  For one thing, I have three personal trees, divided up into my father’s ancestors, my maternal grandmother’s ancestors, and my maternal grandfather’s ancestors. So if I connected any single tree, it would leave out large chunks of my ancestors. If I elected not to choose a particular tree to connect, Ancestry would give a person viewing my DNA connection the option to choose among my trees - but I am fairly certain when some people saw my name in the list with “No family tree” indicated, they would think that I had no tree to connect and would not even click on my name.  Or, they might click on my trees and see my husband’s trees listed as well, which might be confusing.  Better to have something which covers all of my lines, if not necessarily in extensive, documented detail.

So, I created a pedigree tree.  During my first couple of years of research, I looked into what kind of research had been done previously on various lines, took notes, and could reconstruct a pedigree based on this research and some original discoveries of my own.  There were roughly four different categories of research in terms of quality of the material that was online:

  1. Some connections were obviously bogus and I rejected them outright.  
  2. In other cases, there was well documented research online, sometimes including extensive databases that were regularly updated and even discussions among different researchers who hammered out various arguments for and against making specific connections. Many of the supporting documents had been put online. I did not deem it necessary to duplicate these efforts and instead focused on family lines that had not really been researched to a significant extent. 
  3. For other lines, the material that was online was not comprehensive or conclusive, but did seem to point in a particular direction.  
  4. And in the last category, the connections were plausible enough, but I could not find compelling arguments or evidence to nail down the connection.

So now it is true confession time:  my pedigree tree on Ancestry includes some connections from these last two categories.  Before I connected this tree to my DNA results, for the most part I had to search out connections on my own, trying to remember in my head established and tentative family connections.  

[Side note on why I am focusing on Ancestry DNA:  right now it is the major source of my DNA “matches”; I also participate in Family Tree DNA, but so far only have three matches (out of a total of 437 so far) where I can definitely identify a connection, though I hope this will increase as I contact more people. I just received my kit from 23 and Me, but from what I have heard it also has a low rate of responses to inquiries.  From Ancestry, however, I already have a list of 43 connections of interest (for 39 of which I can directly connect family lines) plus a few additional “maybes,” and of course Ancestry is just getting into the game - so far it appears that new matches are added every week, compared to something like twice a month for Family Tree DNA.  Ancestry itself will show the connection if it can, but it’s often possible to figure it out yourself.]

There is, of course, a major fly in the ointment:  at this point Ancestry is not including the actual genetic information, i.e., the precise chromosomal location of the matches.  And this eliminates the main tool for verifying and correlating matches.  My message to Ancestry:  Ancestry, your autosomal matching would be the most useful in the field if you included this information.

The other people I would send a message to would be Ancestry subscribers who are participating in the DNA testing:  if you do not have a tree on Ancestry, create one.  If your tree is private, either make it public or create a separate pedigree tree as I have for connection to match results. If there is “sketchy” information in your connected tree, by all means make that clear when you correspond with your matches.

Below are some of the findings I have from Ancestry DNA (for readers not related to me, you can skip straight to the end past these specifics, but just a glance will tell you that there is some substantial evidence here).  What is interesting (and persuasive) are the consistency of many of the matches:  I can already see trends - multiple matches along certain lines - which confirms the fact that we do not inherit genetic material equally from all ancestors, and may not share genetic material, or at least significant genetic material, with many of our distant ancestors.  Obviously, it is easier to make connections for well-researched lines than for those that are not as well researched, and this makes me wonder about the family lines where I have significant brick walls:  if I could correlate the family lines of my matches, would I turn up some major, multiple overlaps that would give me a hint as to where some of my connections actually are?  (Another idea for Ancestry:  Could you make such a tool available for Ancestry DNA participants?  Just a thought.)

1.  Norman-Monk:  One connection through my great-great grandparents Joseph Madison Carroll Norman and Rebecca Monk.  This is exciting because JMC Norman had three wives, and I am mostly in touch with descendants of his other two wives, so to get an actual connection to one of Rebecca’s descendants is wonderful.  Furthermore, both the Norman and Monk lines get confirmation:

Norman-Read:  This is a 6g-level connection and is significant to me because it goes beyond the 5g-level where, for me at least, the documentary trail seemed a little bit lacking - the will of 5g-grandfather James Norman mentions son Joseph Norman, but I was not sure how he was identified as Joseph Madison Carroll Norman (the older - my 4g-grandfather, the grandfather of my 2g-grandfather with the same name).

Norman-Courtney:  This, at the level of my 7g-grandparents, is a big payoff, and I not only have three connections at this level (plus another possible, but not established connection), I have an independent connection at the next level up for Courtneys (Courtney-Jenkins, the parents of my 7g-grandmother, Frances Courtney).

Monk-unknown:  There are two purely Monk connections, because I do not know or am not sure of the maternal side in each case:  one for my 3g-grandfather Silas Monk, whose wife may have been a Nancy Dunn, and one for my 6g-grandfather Willis Monk, whose wife is not known.  

And from there it gets more interesting and goes into some of those uncharted (for me) waters:

Monk-Hodges:  This one goes to the 7g level.

(Monk-Pool-)Bullock-Hood:  This is an independent connection to my 6g-grandparents, David Bullock and Elizabeth Hood (my Bullock connection comes through a Monk-Pool connection, but the match has no Monk connection), and there is another possible separate Bullock connection.

2.  Brinlee-McKinney:  Two connections at the 2g-level (my great-great-grandparents Hiram Brinlee Senior and Elizabeth Ann McKinney).

McKinney-McClure:  An independent McKinney connection at the 3g level (Daniel McKinney and Margaret McClure) and one at the 4g-level (McKinney-Blatchley) (I also have a McKinney connection at this same level through Family Tree DNA), plus an independent Blatchley connection (at the 8g level - definitely some of that “iffy” territory for me).

3.  Poole-Manning:  one connection at the 4g level (I also have a match at this level on FTDNA).  Next along this line:

(Poole-Manning-)Mabry-Bradley:  One connection (and another possible) at the 6g level. 

4.  Johnson-Moorman and Johnson-Massie:  These would be at the 6g and 7g level.  There are a number of connections here, plus other possibles.  Not surprising that this line, while distant, has a strong showing - these families, plus the Clarks (for whom I have a 7g and possibly an 8g connection) intermarried with incredible (and confusing) frequency.  I always know when I see a man named Moorman Johnson (or Clark Johnson or a family full of Micajahs and Bollings) that there is a family connection.  One of my contacts has a Benjamin Johnson who married a Margery Massie; this Benjamin Johnson was originally posited as a relative of my Johnsons (I think he may be a brother of my 6g-grandfather) and subsequent DNA testing has shown that this is so, plus his Massie wife is almost certain to be related to my Massies.  One of my matches has so many Clarks, Johnsons, Massies, and Clarks in his line that I cannot even figure out all of our connections.

5.  Hamilton line:  One connection at the 6g (Hamilton-Kincaid) level and two at the 7g level (Hamilton-Adams).

6.  The Tarrant line:  I am almost 100% certain that my great-great grandmother Emily Tarrant came from the Tarrants of Greenville, South Carolina, all of whom were descended from a Leonard Tarrant (and one of my matches is descended from this Leonard Tarrant), but I do not know precisely who her parents were.  However, I have been pulling up matches (three so far) with people who have connections to the Dalton families that married into these Tarrants - specifically, Leonard Tarrant’s son Benjamin married a Mary Dalton, and their sons Robert and Wyatt also married a pair of Dalton sisters (apparently cousins).  There is also a Tankersley connection (which may not the actual point of connection with this particular DNA match, as it points to a different son of Leonard Tarrant) and one or two tentative Terry connections (which might point to Emily’s father being one of the two sons of Benjamin Tarrant who married Dalton girls, since their father, Solomon Dalton, married a Terry - yes, I know, this is making my head hurt, too.)   

Obviously, the absence of information on the chromosomal location of my matches, as well as the big gaps in my family lines left by some major brickwalls (two great-grandparents are missing and 16 great-great-great grandparents are missing) means that I may be misidentifying connections.  However, I believe that the information above indicates some real trends, and will be checking all my new connections as well as contacting matches without Ancestry trees (or with private trees) to try to support these findings.

I will be amending and extending this pedigree tree as I turn up genetic and documentary evidence for and against the various connections that it indicates, and I hope that one day it will rise above the level of a “junk” tree, but in the meantime, I am definitely finding that it is a useful (not to say absolutely reliable) tool.


  1. Nicely done. What some may call a "junk" tree, others of us call early research leads. After being hacked, I am rebuilding our TNG website. I have uploaded my full RootsMagic GEDCOM. It has unconfirmed connections and people who have been pulled from undocumented online trees. This is my research locker, filled with people and connections who are waiting to be either confirmed or rejected. Of course, that probably gives me 10 or 15 years of work to do.

    1. Thanks for the vote of confidence, Bart! I do hope to "fix the tree up" in coming years, but I am thinking even now that I should have had something like this long ago, and I like your description of a "research locker."

  2. Nicely done. What some may call a "junk" tree, others of us call early research leads. After being hacked, I am rebuilding our TNG website. I have uploaded my full RootsMagic GEDCOM. It has unconfirmed connections and people who have been pulled from undocumented online trees. This is my research locker, filled with people and connections who are waiting to be either confirmed or rejected. Of course, that probably gives me 10 or 15 years of work to do.

  3. I admire your work and research in using your DNA results. I wish I were as motivated, but I'd be using only my dad's and aunt's results, so my entire maternal line is missing there, as I have not taken a test.

    I think it will be very cool to watch this unfold for you.

    Dee at Shakin' the Family Tree

    1. Thanks, Dee! I think having DNA tests on one side is very useful (especially when more than one person has tested), as you can focus just on those lines; if you eventually take the test, you will have a good instrument for dividing up your matches. The process may look pretty tedious, but I have to admit to being a bit of a statistics geek.

  4. Greta, I am doing the same with my "DNA tree." I have DNA connections that match branches that I am sure belong in my tree- I just don't have enough documentation to support the branches. Glad to read that you have been lead down the same path and are handling it this way also.

    1. Hi Jody - Thank you for your kind comments. It's funny - I'd bet before getting into the DNA side of genealogy neither one of us would have considered doing something like this. Even if the particular family where we connect with our matches is not quite right, we can hope that a combination of research and further DNA results will point us in the right direction.

  5. Good luck, Greta! I look forward to reading about this in the future!

    1. Thanks, Lori - I can't wait to see how the whole DNA thing develops!

  6. What a great post - and I love Bart's terminology "research locker". I always hesitate to even talk about my tree on Ancestry, because it's a mis-mash of things...things that are possibilities, thing that I want to follow up on etc. I'm much more selective about what goes into by database, but for me Ancestry is my playground.

    Good luck with your DNA searching - I really want to understand this better, but haven't yet done much with it. It was very helpful to read what you are doing to pursue these leads.

    1. Thank you, Diana - I also found Bart's "research locker" to be a very apt and useful term. And I agree that it's important for one of our "tree" databases - Ancestry, personal program, whatever - to be an area where we can include our own speculations and even those of other researchers. Hope I can continue to "work" the DNA angle to help my research, because some of these lines can definitely use the help!

  7. I never thought to do this and have only checked out a few of the matches on the Ancestry DNA. I really wasn't sure where to start so your information is really helpful. Thanks

  8. While I haven't yet done this for DNA results I have had great success using temporary "research" trees on over the past several months.

    I'll typically create a research tree to sketch out a new family member or branch that I'm trying to prove/disprove. Using the leaf hints I gather additional information that I may not already have and connections to trees created by other members. I can then aggregate all this data to see whether I can prove (at least up to Gen. Proof. Stds.) that this is a valid connection (or not).

    When I'm done I'll either add the new members and data to my tree (if proven) or discard the old tree if a relation was inconclusive.

  9. Hi Greta! You've just been nominated for the Wonderful Team Member Readership Award! Thank you for taking the time to read my blog and post comments. You can find the post here:

  10. This comment has been removed by a blog administrator.