How does DNA testing confirm family branches?

Question

I know very little about how DNA testing and confirmation works, although I have read over the help pages. I am curious if the possibility exists to resolve questions that are not able to be resolved readily on paper:

I have an ancestor (as several of us here do), John Crawford (https://www.wikitree.com/wiki/Crawford-451) whose paper trail is, admittedly, lacking in clarity. Some believe that he is descended from the Crawford line from Scotland that can be traced back through landed and titled folks to the Crawfordjohn branch of the family. 

If I have relatives who have taken DNA testing, and/or I take a test myself, and we have paper trails that confirm ancestry to John, then good, that confirms we come from his branch of the family. Yes, I know, "confirms", work with me here.

If family members who are descended from the branches that stayed in Scotland or emigrated elsewhere have also taken DNA tests, is it possible to confirm that John's line converges with theirs and therefore he is in fact descended from that family, although not documented precisely? Or is that way too far back in time to be reliably solving this mystery?

To muddy the waters, it appears his son David (https://www.wikitree.com/wiki/Crawford-335) married Jane A Crawford (https://www.wikitree.com/wiki/Crawford-437), and they both may have had the same 5G-grandfather, if some trees are to be believed. 

While it would be nice to find proof of relationships for John, for me it would suffice to say "well, we don't have a paper trail but we can say the DNA evidence shows he was indeed a member of that branch of the family". (Until the mysterious Scottish Dead Loch Scrolls show up with written confirmation attested to by James I, co-signed by Cromwell and witnessed by Nessie)

Answer 1

Hi Jonathan:

You are actually asking a specific question about using DNA to prove your Crawford ancestry.

Doing this will require a Y-DNA test (the Y-chromosome is passed down strictly from father to son, so testing the DNA in this chromosome for mutations establishes who you are related to in your paternal line - that is, the line carrying the Crawford name). Basically, what Y-DNA analysis does is to place you in a particular Y-haplogroup. The broad groupings (such as I, Q, and R) diverged tens of thousands of years ago; the finer-scale groups (which require more expensive analysis to determine) diverged hundreds to a thousand years ago. If you are in a finer-scale group that also contains all known members of the Crawfordjohn branch and does not contain known members of other branches, then you have established your goal.

There is actually a very active Crawford Y-DNA project at FamilytreeDNA (also known as FTDNA) associated with the Clan Crawford Association. I suggest you check out their pages at https://clancrawfordassoc.org/dna and https://www.familytreedna.com/groups/crawford/about.

I cannot comment on their results, as most of the analysis (for privacy reasons) is available only to members of the Clan Crawford Association. So I don't know whether they have actually established yet which haplogroup the Crawfordjohn branch belongs to. However, those are the people you should talk to about the testing you should do.

Comments

Let me add that, looking over the DNA results page at FTDNA, 5 descendants of your ancestor John Crawford (ancestor of the Virginia immigrant David Crawford) have tested to be in haplogroup I-BY3099. This is a subgroup of the broad haplogroup I2. Out of the 600+ other Crawfords in the analysis, only two have tested in haplogroup I2. The main haplogroup for Crawfords is haplogroup I1. Haplogroup I2 split off from haplogroup I1 sometime around 20,000 to 25,000 years ago (see https://www.eupedia.com/europe/Haplogroup_I2_Y-DNA.shtml). What this means is that your ancestor John Crawford is not related in the paternal line (at least within the last 20,000 years or so) to all the other Crawfords in Scotland, Ireland, and elsewhere.

---

And yet the locations of the five I2-01 that are shared at least are in Kilbirnie and Lanarkshire. (https://www.familytreedna.com/public/crawford?iframe=ymap)

There's a story in there somewhere, for sure. Thanks!

---

In the absence of other materials, I'm going to go with "we were the originals, 20,000 years before you, C-H Geschwind said so"...ha!

---

I need to put here for other relatives that the analysis C. H. Geschwind did here is possibly for [[Crawford-18624|John  (Crawford) Craufurd (abt.1600-abt.1629)]], not for [[Crawford-451|John Crawford (abt.1600-1676)]]. 

It is unclear without contacting the test takers themselves to ascertain paper trails whether they have accurately identified the ancestor in question, as Crawford-451 is often conflated with Crawford-18624, but thanks to the good work of Wikitreers here, we now know them to be different individuals. 

See also https://www.wikitree.com/g2g/1027367/question-about-john-crawford-1600-1676-ayrshire-scotland where Henry states that his haplogroup is indeed I-CTS6433, which is found within four locations on the clan crawford results, including those for Crawford-18624. 

Lastly, see Space:John_Crawford_of_Ayrshire%2C_Scotland_and_Virginia%2C_United_States where I am attempting to stitch together these various threads. 

Answer 2

My knowledge of DNA is limited at best, but my understanding is this...

DNA tracks biology. You get approximately 50% of your father's DNA, and the rest from your mother. They in turn have the same from each of their parents. So your portion of DNA from any one parent could be anywhere between 0 and 100% from their parents (though the extreme endpoints are not likely -- we usually guess an average of 50% -- the point is, it won't be an equal share between them).

So when you match a cousin, you can compare the matching segments and somewhat be able to determine which common ancestor those matching segments come from. (It helps to have additional matching cousins in order to triangulate to the shared common ancestor)

This is where the paper trail comes in -- assuming its accurate. Your DNA matches can help confirm its accuracy, or indicate where errors might be.

Of course, Autosomal DNA can only be reliable back to about 6-8 generations. Beyond that, the matching segments become too small and muddled.

Y-DNA can trace much further back along male lines. And since those males usually share the same Surname, its easier to verify paper trails.

mtDNA is similar to Y-DNA in that it can go back about as far, but since mothers Surnames often change with each generation, its much more difficult to verify paper trails.

my thoughts here are mostly oversimplifications.
hope it helps.