Monday, July 1, 2019

ThruLines No Longer in Beta

It's official, ThruLines is here to stay and DNA Circles have been permanently retired. If you're upset by this, check out my article on why ThruLines really is an improvement.

More than that, they've rolled out a few improvements to ThruLines and the match review page. Nothing major, but small things can have a huge impact on your workflow and several of these new items will do just that.

In ThruLines/Common Ancestors:

When viewing pathways with a DNA match to a common or potential common ancestor, clicking on individuals now first brings up a side bar on the right that lists "Relationship Records", which DNA matches have that individual in their trees, and which other trees (not DNA matches) have them in their trees. You can then click on any of those trees and basic data will appear in the same side bar so you don't even have to leave the DNA pages (though of course if you want to bring up someone's full tree you still can by clicking from the side bar). This really helps to streamline your workflow by not having to open as many new pages.



And let's go back to this new "Relationship Records" feature. In the side bar, it will list what records can be found which identify someone as the parent/child of someone else, helping to prove ThruLines pathways may be accurate and not 100% based on tree data. Of course in the example above the only Relationship Record found was for Find A Grave, which is not necessarily reliable - Find A Grave doesn't exact include source citations and often people get linked to the wrong families, etc. So be careful even with these Relationship Records, but they should be useful in general.

Another improvement you might have noticed is the "evaluate" tag shown on anyone who isn't in your tree already. This helps point out (particularly to newcomers) that ThruLines is only a suggestion and you still need to verify (or disprove) it before accepting it as fact.

On the match review page:

We can now view a 10 generation pedigree preview from the match review page like we used to be able to do before all these changes. Granted, you now have to click on "expand tree" which opens a popup box, and I believe this is only available to subscribers, but it's better than having to open a match's full tree in a whole new tab. And while non-subscribers can only view the 5 generation pedigree preview (so I've heard), that is better than the nothing they had before. So this should be an improvement for subscribers, and a big improvement for non-subscribers.


The preview of basic data on an individual in a match's pedigree is also back. When you click on someone in the pedigree preview, instead of opening their full details in the full tree in a new tab, we see the return of the popup box with vital data: birth, death, parents, spouse, and children. These are the basic things we need to more quickly identify a DNA match without getting cluttered with everything else that might appear on an ancestor's profile in their full tree.


But it's more than just the return of what we had before. If you notice in the expanded pedigree preview above, shared surnames are now highlight in green, something we've never had before! Of course, shared surname doesn't necessarily mean that's the branch which your relationship is from but the incorporation of shared surnames into the pedigree preview could be really beneficial for quicker reference.

There are still a few items I have complaints about, but they are mostly regarding the match review page, not ThruLines. It may be worth noting that the match review page, the match list, and the user profile are all still in beta, it's only ThruLines which has been fully rolled out now.

My main issues are:

1. We are we no longer able to select a tree when someone has an unlinked tree. We can click on it, but it opens the full tree view on another page - in the past, we could select an unlinked tree and instead of opening a new page, it would show the surname list, pedigree preview, and map/locations for that tree as though it was a linked tree, and then we could even toggle among other trees if the user had more than one. This was much more useful and without it, it's harder to compare surnames and locations and find a connection with an unlinked tree.

2. The loss of the Shared Locations list. The map is there but it would be much easier to see a list than have to click around the map. I never really used the map in the past, and found the locations list much easier to use.

3. There should be a distinction between "Common Ancestors" that appear in both my tree and my match's tree, versus ThruLines/Common Ancestors being found through combining multiple other trees together, which is much more speculative. If I have the same ancestor in my tree that a DNA match does, that should take priority and be a notable distinction from other ThruLines.

I guess with ThruLines coming out of Beta, this is unlikely to be implemented, but a girl can dream and hopefully with enough feedback, my other two complaints regarding the match review page might change.

Wednesday, May 22, 2019

Why ThruLines is an Improvement (Really!)

I've been hearing a lot of negativity about AncestryDNA's new feature ThruLines (aka Common Ancestors when viewing your match list - they are the same thing, just different ways to view them), which seems to be replacing both DNA Circles and Shared Ancestor Hints. It works by looking for a tree connection between you and a DNA match through other family trees, not just yours and your match's trees.

Naturally, this opens the system and connections up to a lot of room for error. We all know how inaccurate some trees out there can be, how those errors can get copied by the dozens, and I'm not going to pretend that this isn't an issue. Anytime you're working with trees, you absolutely must verify the information yourself before you accept it as fact.

But that doesn't mean ThruLines can't be useful. I've been working with them for a few weeks now, and I have found that the majority of them can be verified with a little work. Some of them I haven't been able to verify at all and remain unknowns, and others I have actually proven wrong, but the amount of them which I've been able to confirm far exceeds the old Shared Ancestors Hints which were limited to looking at your tree and your match's tree only. ThruLines is finding legit shared ancestors that neither the system or myself doing the work manually could have ever found before. Take the following example.


As you can see, Sally has a family tree with only 3 people in it and on top of that, it's a private tree. In the past, this would be a complete dead end. Unless I contacted her and she responded to tell me more about her tree or at least give me access (which I likely wouldn't do for every match at this distance), I would have completely written off this match in the past.

But now, ThruLines has found a common ancestor! Just by adding those mere 3 people to her tree and making her tree searchable even if it's private, Sally has allowed ThruLines to find a connection between us. Of course, I first have to make sure ThruLines isn't leading me astray, so let's look at the connection.


Part of what you're seeing here is the benefit of researching as far down the descendant lines of your ancestor's siblings as possible. When I first looked at our ThruLines relationship, I had only researched down to Samuel Gross, and Sally's tree only went back to her grandparent. The two generations in between were being filled in by ThruLines finding them in other trees. Of course, to verify the relationship, I had to do my own research and fill in the missing generations, but this was much easier since I'd already gotten a head start. I researched all the way down to Sally's grandfather, Robert. Since her tree was private, Robert was originally shown as private, but you can actually click on deceased people in private trees and get basic details like their name and birth, just like you would from clicking on results of a private tree in the search engine results (see below). If they are still living, you get nothing, as it should be - living people's privacy is always totally protected (also see below).



In the past, if a tree was private, we got nothing unless the tree owner chose to share info or an invite with us. But with ThruLines, we can now see which common ancestor we share and get basic data on each generation even from private trees, enough that we can then go and research the pathway ourselves and verify it, which is exactly what happened here.

ThruLines isn't perfect, of course. It's only as reliable as the trees it's using, and I have found errors in the trees it uses on occasion. Additionally, the system itself is not infallible and I have come across the occasional case where ThruLines is assuming two different people are the same person. In those cases, the error is with the system, not the trees it's using.

But we are genealogists - fact checking and verifying data is what we do. So let's do it! You might find ThruLines just as beneficial as I have so far. It has found common ancestors with people that the old system wouldn't have, and with people I never found have by searching manually on my own.

Monday, May 6, 2019

Are they really a 4th-6th cousin?

At all DNA companies, matching with other testers includes an estimated relationship (usually a range) based on how much DNA you share. The key word is "estimated". There's plenty of examples of finding relatives whose relationship to you actually winds up being outside the estimated range.

AncestryDNA has the option to click on how much DNA you share (when you're on the match review page - click the little "i" icon if you're on the match list page) where previously, it just gave us generic info about that estimated relationship degree. Now (and this may only be available on the new beta match list), it gives us specific probabilities of each possible relationship based on the exact amount of DNA you share. What I found is that sometimes, even their statistical probabilities don't stick within the estimated relationship range.

Above left is an example. This is with a confirmed 3rd cousin of my mom's, an estimated 4th-6th cousin. Again, it's not unusual that the actual relationship is outside the estimated range, but AncestryDNA's predicted relationship range doesn't even seem to be following their own statistical probabilities. As you can see, 65 cM shared has the highest probability (32%) of being 3rd cousins (or 2C2R, Half 2C1R, or Half 1C3R), which isn't even within their predicted 4th-6th cousin range. No prediction system is going to be perfect (because there's a lot of overlap between different relationships so no estimate will be exact), so I'm not saying anything is wrong with the system, I'm just pointing out that this new statistical probability option will tell you a lot more than the predicted relationship range, so don't overlook it.

Of course, this doesn't mean the actual relationship will always be the one with the highest probability. Lower down on my mom's list is a known 4th cousin sharing 57 cM, and the highest probability is still 3rd cousins at 28%. 4th cousins only have a 16% probability, but this match falls into that 16%.

These probabilities will especially be great for unknown close relations popping up, or telling the difference between full and half siblings. My grandfather's half niece, for example, has an estimated range of 1st-2nd cousin, but in the probabilities, it lists all the possible relationships, including half niece/nephew at 96% probability. In fact, at the amount of DNA they share, it doesn't even list 2nd cousins in the probabilities, ruling that out completely even though it's a part of the original predicted range.

DNA Painter offers the same type of probabilities, but collected from multiple companies, not just AncestryDNA data. It's a great tool for matches from companies who don't provide these probabilities, but now that AncestryDNA does, their internal probabilities are likely a better option for your AncestryDNA matches, since the data is specific to them.

Friday, April 26, 2019

More Changes from AncestryDNA

Yesterday, I clicked on a match to discover the old "Match Review" page and the newer "Compare" page had been combined into one. I kind of figured that's where they were headed with the changes they'd made to the beta match list, but I was pleased to see it finally happened.

Previously, there were certain features and tools missing from the newer "Compare" page, but the new "Common Ancestors" wasn't featured on the older "Match Review" page meaning you often had to open both pages to make full use of your matches. Additionally, there was no way to add a match to a group from either page, you could only do so from the match list. All of that has been resolved now, and they also reduced the size of the profile pictures, which previously took up almost the whole page on smaller screens - a complete waste of space that only forced us to scroll down more to see any useful information. These were some of my top complaints and now they're all fixed.

As you can see from the screenshot above, everything you need to make use of your matches (well, everything that AncestryDNA current offers) is on the single new page. You can add and edit notes, you can add the match to a group, you can see how much DNA you share without even having to click on anything, the pedigree preview is there (with the option to open their full tree), and Common Ancestors will show on the left if there are any. There's also the option to click on a built in tab to see your compared ethnicity and one for Shared Matches.

Where are the surname lists and location map? Don't worry, they're just a little further down the page (screenshot to the right). The only disappointing thing is that these no longer show all the way back to 10 generations, and the pedigree preview no longer shows back to 7 generations. They are now all limited to a mere 5 generations, which seems much too restrictive in my opinion. That means they only include back to your 3rd great grandparents, which means they will only apply to as distant as 4th cousins! I think we can all agree that there are many, many more distant yet legitimately identical by descent matches than that. Even ThruLines/Common Ancestors goes back further than that (5th great grandparents, or 7 generations). What this means is if there's no ThruLines/Common Ancestor found for a match, we're going to have to open up more pages to explore most matches deeper for a common ancestor, and that seems counter-productive to an efficient workflow. It almost feels as though AncestryDNA is trying to force us to be more reliant on ThruLines than anything else, which would be a big mistake considering ThruLines will only ever be as reliable as the trees it's using.

If you agree with these complaints and/or have any others of your own, remember to keep sending feedback to AncestryDNA through the "Prodvide Feedback" option in the bottom right corner of your match list page!

Monday, March 4, 2019

New Tools from AncestryDNA

ThruLines from the DNA homepage,
you can still go back to DNA Circles
Last week, coinciding with RootsTech, both AncestryDNA and MyHeritage announced several new tools and features to help you get the most out of your DNA matches. AncestryDNA rolled out ThruLines, which it seems will be replacing DNA Circles, and New Ancestor Discoveries, although for now you can still access the latter two. They've also completely revamped our match list, though this is only available in beta mode if you enable it in Ancestry Labs available at https://www.ancestry.com/BETA (you can disable it in the same place if you wish).

ThruLines works by automating sort of what I've been advocating people do manually for years: build on your match's trees to find a connection. Instead of only looking at your tree and the trees of your matches (like Shared Ancestor Hints and DNA Circles did), ThruLines looks to other trees outside your DNA matches to make the connection between you and your matches by building on both trees (much like how the We're Related app worked). As with anything tree-related, be careful about the connections it makes - I noticed one potential ancestor showing up because the system decided my ancestor Kate White was the same as someone called Katherine Weiss, which is German for White, but my Kate White is Scots-Irish, not German. And of course even putting aside the system making connection errors, the research other users do on their trees may be wrong.

ThruLines details/pathway
So do your own research and confirm or deny the relationship, but ThruLines is a great way to get you started and bring potential common ancestors to your attention so you're not just aimlessly building your matches trees, unsure of what to look for. The best thing about ThruLines? Unlike with Shared Ancestor Hints, it will actually show you the common ancestor (and pathway to them) even for private trees! Hooray, no more seeing that you have a Shared Ancestor Hint and messaging people to ask who it's for and never getting a response! Of course, private trees are still private in that you can't view the rest of the tree without an invite, but at least you can see your common ancestor and the pathway by clicking on the "private" ancestors in each generation of the pathway (it will take you to a page that lists their name and birth details only, but this is generally enough to confirm or deny the pathway). In the screenshot to the right you can see the common ancestor found with one of my matches through ThruLines. Clicking the numbers will expand the pathway and show each generation. Ancestors in dotted lines indicate ones which ThruLine has found through other trees - note that it names the tree owner and if you have a subscription, you can click on that ancestor and see their tree if it's public. One thing I don't quite understand about ThruLines is that it doesn't group married couples together - above you can see only one ancestor, Elisha Mills, is listed. His wife is also included as a common ancestor but listed separately. Previously, Shared Ancestor Hints included spouses as a part of the same hint.


If you've enabled the beta match list, you'll see some quite drastic changes (shown above). Your match list is now on an infinite scroll, color/custom groupings have been added, and you have several new sorting and filtering options. I don't love the infinite scroll - although the additional sorting and filtering options does make an infinite scroll less cumbersome, it is still cumbersome when you consider you have literally thousands of matches. That's a lot of data for an infinite scroll and basically assures you'll never reach the end of the list even when using most filtering options to narrow the numbers down. Additionally, if your browser crashes or you have to shut it down for any reason, you will lose your spot in the infinite scroll. With pagination, you can just note the page you were on and jump back to it, but that's not possible with infinite scrolling. While you may be able to find the spot you left, it's still going to take time to scroll, scroll, scroll down until you finally reach it. Infinite scrolling is not very functional and I really hope AncestryDNA's developers reconsider it.

Custom groups offer 24 different colors (shown above and right), so you can now group your matches however you want, such as a different group for each of your sixteen 2nd great grandparent's branches, and then a few more for other options (like maybe a group for a certain location, or an unknown cluster you've noticed, or matches you've identified a MRCA with, etc). You can add one match to as many different groups you'd like and each group gets a name that appears if you hover over the colored dot. This may eventually replace what I've been using the emojis for in the notes section (you can see they are still in my screenshot above), but not just yet since at the moment, groups are only available on the match list page, not the match detail page(s). So I still need emojis as a visual reference on the match detail page. I imagine this may change in time. The new match list is, after all, an opt-in beta mode and I'm sure there are tweaks and additions to come. If groups get added to the match detail page(s) then emojis might become unnecessary, unless you're using more than 24 of them (I am not). The benefit of groups over emojis is that you can filter by groups. And although Chrome extension MedBetterDNA allows you to filter by hashtags, it was cumbersome since it only hid matches without the selected hashtag and did not condense them - so a match with the selected hashtag on page 50 of your matches would still be on page 50. That is no longer a problem with AncestryDNA's groups (or infinite scroll, although that has it's own issues addressed above), however, since groups are limited to only 24 options, whereas hashtags are unlimited (there's a limit of how many you can search at a time in MedBetterDNA, but you can use as many as you'd like), don't get rid of those hashtags either. I still have all identified surnames hashtagged (and will continue to add them) because groups being limited means I can only use them to identify branches, not individual surnames, as there would be too many surnames to use with groups. So groups will likely replace emojis eventually, with added benefit of filtering groups, but it will not replace hashtags used with MedBetterDNA. Additionally, MedBetterDNA's essential feature of being able to always show notes means this extension isn't going to become obsolete. It's functionality has been influenced by the beta updates, but the extension's developer expects to update it to work with the new match list once the beta goes public (likely a few weeks after the new match list gets pushed out to everyone).

Other default "groups" (shown above/right) that Ancestry provides allow you to only show 4th cousins or closer, only show distant matches, or hidden matches, and as before, only show new matches, starred matches (colored/custom groups being separate from starred matches), or matches with your mother and/or father (if they have also tested). In fact, it now identifies "Mother's Side" or "Father's Side" in the far right column if your parents have tested (shown above). Note this only includes estimated 4th cousins or closer, so if you want to include more distant matches with your parents, you'd have to create a group for it (which is exactly what I've done).

Under filtering, a separate drop down menu (shown left), you can view matches with a "Common Ancestor". This appears to be replacing Shared Ancestor Hints, and it will include speculative ThruLines, so it's basically just another way to view your ThruLines (viewing them by individual match rather than by ancestor, which is what you see when you click on ThruLines from your homepage). The major downside to this, and I have already sent feedback to AncestryDNA on it, is the fact that there is no way to view only matches which have an actual common ancestor in both our trees versus those which are much more speculative because it's finding connections through other trees. There should be a way to separate them, as there is a noteworthy distinction between them. If you find this problematic as well, please make sure to send them feedback in the lower right corner of your browser while in beta mode (shown below/right).

Additional filtering includes "matches you haven't viewed". In the past, there was a "new" button which showed all matches you hadn't viewed yet and you could sort them by relationship or date. This is now split up so you can view new matches by date (including ones you've viewed) from the "Groups" drop down menu, or view matches you haven't looked at yet which is sorted by relationship from the "filter by" drop down menu. Personally, I use the latter the most, because I've made a point of looking at each estimated 4th cousin or closer and trying to figure out how we're related, so now when new matches come in, I periodically just look at new estimated 4th cousins or closer and examine them to stay on top of it.

Among the new filtering options include viewing matches you've messaged, matches with notes (very useful!), and matches with private, public, or unlinked trees, the latter being something we've all been asking for for years! Finally!

Best of all, you can combine options between the "Group" dropdown and the "Filter by" drop down menu. So if you want to see matches who you've you've sent a message to and labelled on your Smith branch, for example, you can do that.

Clicking the circled items opens the old "view match" page with the pedigree preview, etc

All of these new features and tools could be hugely beneficial, but my main disappointment is in the organization of much of it. In the past, clicking on a match opened the "view match" page we were all used to, including all the most useful features: pedigree preview, surname list, shared matches, map/locations, notes, and the option to message, star, trash, or mark as new. Somewhat recently, there was also added a "compare" button that took us to a different page which showed the comparison of our ethnicity reports, which was very interesting, but typically not very useful. Now with the update, clicking on a match's username takes you to that "compare" page instead, which does include shared matches and potential common ancestors, but does not include the ability to add or edit notes! It also won't show notes of the shared matches listed. Fear not, the old "view match" page with the pedigree preview and other features is still available: from the "compare" page, you can click "pedigree tree" or you can go straight to it by clicking on the tree info in the third column of your match list (even if there's no tree or it's private) - see above (edit: this option has now been removed, in order to get to the pedigree preview page, you can now only do so from the "compare" page). The problem is, the old "view match" page no longer includes Shared Ancestor Hints - in order to view them (now called Common Ancestors, or Potential Ancestors as a part of ThruLines) you have to look at the "Compare" page. And in order to view the details of those potential ancestors (part of ThruLines), it will open a third page! On top of all that, groups can only be added from the match list page, meaning in order to use all useful features, you have to switch back and forth among FOUR different pages. I can understand ThruLines being on a separate page because it's very detailed, but I see no reason why they can't combine the "Compare" and "View Match" pages and/or add grouping to those pages. It's ridiculous that the features are split among them, it's going to make my workflow that much less efficient. If you feel the same, please use the feedback option to make sure AncestryDNA know!

If AncestryDNA could resolve the two bolded issues, it would be almost perfect - now all we need is a chromosome browser and we're all set, but apparently that's never going to happen due to "privacy" issues.

Coming soon, I'll detail the new features from MyHeritage.

Thursday, January 24, 2019

24genetics Review

My European countries map from 24genetics
24genetics offer DNA tests for health and ancestry related reports, and also accept raw DNA uploads from other companies for more limited report options. For the uploads, they only offer reports on ancestry, sport, skin, nutrition, and talent/personality. If you buy a test with them, you can also get health and drug response reports, but be aware that each one costs $149 - $199, and if you want all the reports, the "all in one" pack costs a whopping $399. This seems very expensive, and if you're looking for a health report, I wouldn't buy tests through this company, there are less expensive venues. 23andMe costs half this for their ancestry and health, and is on sale often enough you can get it for even less. 23andMe may not include all same reports, but you can then upload your data to other, inexpensive sources like Promethease for more.

Most of the reports from uploaded DNA say they accept data from 23andMe or AncestryDNA, except the ancestry report, which says it will accept DNA from 23andMe, FamilyTreeDNA, AncestryDNA, or MyHeritage. The ancestry test costs $49, and the rest cost $69. You can order a promo pack including ancestry, sport, nutrition, and skin for $99. These also seem expensive for what you get, considering they don't have to run your DNA through a lab, so I only purchased the ancestry report and uploaded my raw DNA data from 23andMe.

The report is sent in an email as a PDF, much like how DNATribes before they discontinued their upload offer. In the PDF, you get an ethnic break down of your DNA first on a continental level called "Global Vision", then a country level, and finally a regional level. Like most ethnicity reports, they say the results date back "hundreds and even thousands of years".

My country results from 24genetics
My continental results (deserving of all caps, apparently) say:

EUROPE 99.20%
ASIA 0.80%

Most likely, the Asian results are just noise. You'll see in the country and regional break down that my Asian results are in a part of Georgia, in the Caucasus area (just above the Middle East), so it's possible this is related to my southern Italian ancestry, but given the small percentage, it may just be noise and not mean anything.

My country results:

Great Britain 33.80%
Italy 30.00%
Austria 17.10%
Greece 8.20%
Netherlands 5.60%
Switzerland 2.30%
Finland 2.20%
Georgia 0.80%

The top two results are very accurate, my family tree is indeed roughly 32% British, and I did in fact inherit about 32% of my DNA from my Italian grandmother. But rest of the results aren't very consistent with my known ancestry. The smaller results could just be noise, and I'm guessing the Austria result is coming from my Germanic ancestry. I'm not sure where the Greek is coming from, since the only Mediterranean ancestry I have is already fully accounted for in my Italy results.

AncestryDNA's PCA chart showing reference panel
populations and their genetic distance to each other
The main thing it's missing is my Norwegian ancestry. Though I had one Norwegian great grandfather, my Scandinavian results often come back lower than the "expected" 12.5% on most ethnicity reports, so perhaps I inherited less from him than expected. I do get a trace amount in Finland, but as you'll see below, it's in an eastern part of Finland, which seems like it would have more in common genetically with Russia than Norway. In fact, AncestryDNA used to group Finland and NW Russia together. Although they now have separated the two groups, their PCA chart (left) shows that the Finnish group has no overlap with Scandinavia (let alone Norway), but does have some minor overlap with the Baltic States (northeast Europe, such as Estonia, Latvia, Lithuania). So I doubt there's any connection to my Norwegian ancestor. Norway usually has more in common genetically with Germanic Europe and Britain, so maybe the Netherlands results are coming from my Norwegian ancestry, but I do also have Dutch ancestry, though much further back on my tree.

Maps are provided for the results, first showing the highlighted countries on a global scale (above right), and then zooming in on Europe (above left). I guess my one small result in West Asia didn't warrant a zoom in.

My regional results
My regional results:

Essex Medieval 32.30%
Tyrol 17.10%
Apulia 15.80%
Lombardy 13.60%
Utrecht 5.40%
Crete 3.20%
Kythera 3.00%
Romandy 2.30%
Finland Karelia 2.20%
Andros 2.00%
Cornwall 1.50%
Georgia Svaneti 0.80%
Treviso 0.60%
Groningen 0.20%

Essex Medieval is an interesting result because it specifies a time period as well as location. I wonder why the other results don't include a time period? I do indeed have ancestry from Essex, although there's a few branches from there, they are all from fairly far back on my tree, in the 1500s. I suppose that's consistent with the Medieval timeline, however, they seem to be attributing almost all my British ancestry to Medieval Essex, yet I have British ancestry from many other locations, some from much more recent time periods too (which should account for more of my DNA).

My European regional map
Tyrol is a region of western Austria which borders Germany and Switzerland, both places where I do have ancestry from, so I'm assuming that's where this result is coming from.

Apulia is an area of southeast Italy. I have no known ancestry from there, but my Italian ancestry is indeed southern/Sicilian, not northern, so I'll give it points for that. However, Lombardy and Treviso are a part of northern Italy, where I have no known ancestry.

Next is Utrecht, a city in the Netherlands. I don't know where in the Netherlands most of my Dutch branches come from, but one of them is actually from Utrecht, and another from Amsterdam. But let's face it, the Netherlands is a pretty small country (though it was once larger), so I'm not really sure how distinct DNA from different parts of it really are from one another - Amsterdam is really not that far from Utrecht (about 27 miles). Although this is a small percentage, my Dutch ancestry is from far back on my tree, so I would expect it to be a small percentage, if it would show up at all. I also get a very small hit (possibly noise) for Groningen in the Netherlands, where I don't have any known ancestry.

My Asian regional map, probably
noise
Now we get into percentages so low they may only be noise. Crete, Kythera, and Andros are all in Greece, which again, I have no known connection to. Maybe some of it is related to my Italian ancestry, but it can't be all of it, or that tips my Italian percentage over the edge of being too much. Romandy is a French speaking part of Switzerland, and I do have both French and Swiss ancestry (again, far back on my tree), so perhaps there's some legitimacy to this. Karelia is an eastern part of Finland I have zero connection to (again, it's unlikely it's from my Norwegian ancestry, and I have no known ancestry in Eastern Europe). Cornwall is the southwestern most part of England, which I have no known ancestry from, but could still be a legitimate part of my British DNA. Svaneti is a historical area of Georgia, which may just be noise or could be coming from my Italian ancestry.

Overall, I feel like my top most results from this company are accurate but it's missing significant locations and the smallest percentages are likely just noise. An interesting assessment, especially the more specific break down that you don't always get from other companies, but as with any ethnicity report, don't take it too literally. Worth the money? Probably not, maybe if it was cheaper.

Tuesday, January 22, 2019

Breaking Down Brick Walls with DNA

I'm going to detail one example of how DNA helped me break down a brick wall in my tree. This may be a somewhat unique case that might not always apply to every situation, but it's still worth detailing so people can get an understanding of how to work with your DNA matches in general.

Hope's baptism confirming her parent's names, but not her
mother's maiden name.
My 4th great grandmother was Hope Gibbs, b. November 3, 1805 in Philadelphia. I knew from her baptism record that her parents names were Caleb and Isabella (maiden name unknown), but I didn't know anything else about them. My grandmother had left notes with the names and some details of Hope's sisters, and indeed, research of some of them not only confirmed those details but also listed Caleb and Isabella as their parents, but still no details on those parents. I even found some DNA matches who were descended from those sisters, but none of those matches knew more about their parents.

I then found a marriage record of a Caleb Gibbs and Isabella Brannin in 1799, in Mt Holly, Burlington, NJ (Ancestry.com wrongly transcribed the location as Philadelphia), but given the difference of location, how could I know it was the right couple I was looking for? Further research on the couple married in 1799 revealed they were Quakers, and Isabella was the daughter of Barzillai Brannin/Branin, and the granddaughter of Michael Brannin/Branin, but I still was unsure about the connection to Hope Gibbs. Were these really her parents?

It took me a while to think of this for some reason, but I eventually decided to look for the surname Brannin in my DNA matches (well, my mom's DNA matches, since she is one generation closer to the ancestors in question). Theoretically, I figured if my Isabella was a Brannin, I (or my mom) would have DNA matches with people descended from her father or grandfather, I just hoped Brannin wasn't so common of a name that I got unrelated hits for it among my matches. It would be much more difficult if the name was something like Smith, and indeed, I had previously tried to search my DNA matches for the Gibbs surname, hoping to find people who descended from a father or grandfather of Caleb, but Gibbs was too common of a name and I was getting results for DNA matches with obviously unrelated Gibbs ancestors due to their location or time period (my DNA relation to these matches obviously lies elsewhere and the Gibbs surname is just a coincidence).


But the results and conclusions of the Brannin surname search were almost instantaneous - I quickly found 10 DNA matches descended from either Barzillai or Michael Brannin (two shown above), which would only be likely if I was descended from or at least related to them genetically. Not all recurring ancestors among your DNA matches will be your ancestors, some may be related to you in other ways, but combined with the existing question of whether my Isabella was Barzillai's daughter, the DNA matches conclude that the two Isabella's were indeed the same. Worth noting is the fact that two of the ten matches shared a segment with my mom of over 14 cM, which has over a 99% chance of being identical by descent, so these are not false positive matches. Also worth noting is the fact that some of the ten matches I found I had to uncover myself because their trees didn't go back far enough, but with a little digging, I expanded on their tree and trace their Brannin line back to Barzillai or Michael. Sometimes, you have to do a little of the legwork yourself.


Quakers are good record keepers, so confirming Isabella's identity was a huge breakthrough for me, and I then spend the next few days building this new branch of my tree, going back several more generations on both her paternal and maternal sides, something I couldn't have been sure of if I hadn't found the DNA connection. Of course, this also confirmed Caleb's identity as the man who married her, but unfortunately the Quaker records didn't go back as far for his ancestry. It at least confirmed his parents as Samuel Gibbs and Mercy (maiden name unknown).

By the way, if you're wondering about the notes showing simultaneously with both DNA matches in the screenshot above, that is with the use of Chrome extension MedBetterDNA. It allows you to always display notes so you can refer to them quicker and easier. It also has a lot of other useful tools, like the the hashtag option, which you'll see I'm also making use of - it lets you search your matches by hashtags in the notes field, so you can more quickly find the matches you're looking for. You may also notice the emoji icons I'm using - the green check mark and the black heart. There's no hidden meaning to the black heart, I'm just using it as a quick visual reference for that particular branch of my tree. Every one of my 2nd great grandparents gets a unique icon for their branch and I was running out of heart colors - that may seem excessive with 16 second great grandparents but it's working for me. The green check mark denotes matches with whom I've identified our most recent common ancestor. As a visual person (I am a photographer, after all), I just find the icons help my brain sort my matches. For non-windows users, you can also make use of flag emojis (the flags don't work on Windows for some reason) to identify matches with a MRCA from a known country.

I hope this helps illustrate how you can use your DNA matches to confirm questionable branches like this. Keep in mind that the further back the ancestors you're looking for are on your tree, the less likely you'll be able to reliably use or find identical by descent DNA matches descended from those ancestors. In this case, we were looking at my mom's 5th or 6th great grandparents, meaning she'd be around 6th or 7th cousins with those DNA matches, which is still within the realms of identical by descent.

Wednesday, January 16, 2019

Genealogy Photos and Copyrights

Unknown women, possibly sisters.
The backs of these are stamped with photog's info,
J. Bronson of Philadelphia, who died 1914
This is a topic I've touched on before and something which has been very well covered by The Legal Genealogist too, but since it still persists in coming up in the genealogy field as greatly misunderstood, I feel the need to go over it thoroughly as well, as both an amateur genealogist and professional photographer. The following only applies to US law, though other countries may have similar laws.

Many people seem to be under the impression that because they inherited a photograph, particularly one of their ancestor, it means they own all rights to the image, and get upset when someone else, usually another descendant or relative of that ancestor, "steals" it and adds it to their own tree, or shares it in some other way, especially without "permission" or at least "giving credit" to them. Sure, it might be a little rude, but before you get outraged and indignant, first be sure that you really understand copyright ownership.

Most family photos are unpublished works, which means (unless otherwise agreed upon, which isn't typical) the copyright is retained by the photographer for the duration of their life plus 70 years after their death. How can a copyright be retained by someone deceased? Just like any property, their heirs inherit it. So unless you or your ancestor took the photograph, you do not own the copyright, and if the photo is still in copyright, you may actually be the one committing copyright infringement by publishing it online without the photographer's or their heir's permission. It isn't exactly fair to get angry at someone for "stealing" a photo you had no right/permission to publish online to begin with.

If the copyright has expired (so it's been 70+ years since the photographer's death), then no one owns it and no one is legally doing anything wrong by copying a photo you shared. Sure, it might be a little rude to do so without a thank you, but when you make an expired copyright image publicly available, you have to assume it's going to get copied, because people are going to assume you put it out there to share it, not just show it off. And as for "giving credit" - I can certainly see the value in identifying the source of who originally scanned and shared the photo, so that any questions about it can be redirected back to you, however, that is called citing a source, not giving credit. (While on the topic of giving credit, it is also worth noting that when an image is still in copyright, giving credit even to the rightful copyright holder does not absolve you from copyright infringement, only plagiarism. Copyright is the right to publish, distribute, alter, etc so if you do not own the copyright and did not get permission to do so, you do not have the right to publish it, even if you "give credit". A lot of photographers today will let it slide as long as you do give credit, because it's free advertising, but that's their prerogative, they still have the right to send you a cease and desist letter if they want to, so it's always better to get permission first.)

Maybe your ancestor did take the photograph. Maybe your ancestor was a professional photographer, or it's just a candid snapshot taken among family with a personal camera - they were available from 1900 onward, so it's very possible. Generally, professional shots are easily distinguishable from family snapshots, at least in history (today, there are lots of amateur hobbyists as good as professionals). Don't be fooled into thinking unprofessional photos aren't copyrighted, they are. The trouble with them is knowing who took it. If you can determine the photographer as your ancestor (pro or not), and that ancestor died less than 70 years ago, then congratulations, you probably do hold the copyright for that photograph. Only catch is, probably so do all the other descendants of that same ancestor, and if one of them was the person who copied the photo you shared, then technically, they are well within their rights to do so.

In conclusion, the only situation where you have any ground to stand on when accusing someone of "stealing" photos you shared of your ancestors is if your ancestor took the photograph, died less than 70 years ago, and the person copying the photo is not also a descendant of that photographer (pro or not) ancestor. I understand even if it's not a copyright violation, it still bothers people when someone doesn't say a simple "thank you", but all I can say to that is if it really bothers you that much, don't put photos online, or make your tree private (note: private trees are not available at FamilySearch or Wikitree). But if you do put them publicly online, and you're not the copyright holder, then you really have no one else to blame but yourself.

The photographs I've included above as examples were stamped on the back (shown right). J. Bronson turns out to be James (aka Jay) Bronson who was born May 23, 1862 and died April 13, 1914. Since this was well over 70 years ago, it means his photographs are assuredly in the public domain, which means I'm not only free to publish them without permission, I also really have no room to complain if someone wants to copy them. I would hope that anyone who does though could cite my blog as the source so anyone with questions about them, or seeking higher quality versions for printing, or anyone who knows who these women are can get in touch with me.

On another note, I have been unable to find Bronson at this particular address of 46 Main Street, Germantown, Philadelphia, Pennsylvania. Although there are many other listings of him on Main Street (aka Germantown Ave), none at number 46. That makes it difficult to narrow down the dates of the photos, but I believe they were taken at the same time, and that the similarity of their dresses suggests they may have been sisters. I'll go into more detail on this in another post.

The Legal Genealogist goes into more details about copyright laws in different situations, but I focused on the most common situation for most genealogists with photos of their ancestors, as it's the most applicable.

Sources:

Monday, January 14, 2019

Dating Old Photographs: Example #4

A cabinet card of an unknown woman. I have a lot of unknowns in my family's collection, and I think many of them were actually friends or distant relatives of my ancestors that my family simply held onto even though they weren't labelled. Given the location and time period of most of them, I think they were probably friends or relatives of my Rorer or Fallows ancestors in the Philadelphia area.

Estimated date: 1892-1893

The first thing I looked for was the photographer. It may not look like anything more than a decorative flourish at first, but that swirly thing in the middle of the bottom margin is the photographer's monogram and if you look closely it's the initials A.P.K.T. City directories tell me this was Albion K.P. Trask and that his photography studio was found at 1210 Chestnut Street in Philadelphia from 1879 to 1893. This narrows it down to almost a decade.

The curly/frizzy bangs (or fringe, if you're British) were popular throughout the 1880s and into the early 1890s as well, so her hairstyle is very consistent with the photographer's studio years.

The dark color of the cabinet card is also telling. These were used from about 1883 to 1895, although they were more expensive, so lighter cards were still dominant at the time. This helps narrow down my time frame, but can I narrow it down further?

The clothing she's wearing is very interesting. From time to time, I see this kind of beaded embroidery on the chest, especially on shiny, satin material like this. It seems to have been popular in the late 1880s and early 1890s. Given the lack of puffed shoulders or sleeves, I would normally think it's more likely to be late 1880s, but I can't completely rule out early 1890s. Just because a certain trend was fashionable during an era doesn't mean it was the only thing people wore.

Even more telling is the back of this cabinet card. The artwork covers nearly the entire back of it, which according to Phototree puts it in the range of 1888-1900. But most notably is the beveled edges. You might be able to see them in the image above, decorated in gold. Phototree tells us that beveled edges were used 1892-1900. We already know this particular photo is not later than 1893, due to the photographer's address, so the beveled edges narrows it down to 1892-1893.

The other elements, though they may not have narrowed it down this far, are still worth noting and knowing for other photographs, and understanding more about the subject. The extra expense of the dark card and the elaborate artwork on the back of the card suggests this photographer was on the pricier end of the scale, and that tells us something about the woman in the photograph too, she obviously wasn't exactly poor. This also helps me narrow down which branch of my tree she came from (even if she was only a friend of that branch).

On another note, the photographer Albion K.P. Trask was born about 1831, and died November 11, 1900, which was well over 70 years ago, meaning that the copyrights of all of his photographs are expired and now in the public domain. This leaves me (and anyone else) free to share and publish his photographs without permission.

Friday, December 7, 2018

A Tip on Photo Dating

Emma Sarah Fallows, c. 1900-1910
More than once, I've come across people who seem to think that if a particular trend was in fashion during a certain era, it means people didn't wear anything else! This is definitely not true. Just like today, there was often more than one trend popular at the time, and not everyone wore the most trending fashion, at least not all the time.

Recently, I saw a group photo (shown below right, used with permission from owner Ashley McQuillen, who asks relatives to contact her by email) including several women wearing the popular light colored shirtwaists with darker skirts, which were popular around 1898-1910, apart from one woman who was wearing a darker colored blouse. An assumption was made by someone that because light shirts were popular at the time, the only exception to this would be if someone was in mourning and therefore wearing black.

Please, don't go down this path of tunnel vision.

The photo, which subsequent discussion inspired this post.
The Oliver family of Cumberland Valley, Bedford County, PA
Back row, left to right: Susan Wertz Oliver, her husband
Patterson Oliver, unknown woman, Francis Morgan Oliver,
his wife Jane Gillum Oliver. Front: unknown woman,
Owen Ash Oliver, Ida Oliver Mock, 2 unknown women,
Hosea Hudson Oliver
Just because lighter blouses were popular at the time doesn't mean the only possible exception would be someone wearing black for mourning. I am not saying the woman wearing darker colors couldn't have been wearing black, or couldn't have been in mourning, just that you can't make that assumption, partly because that wasn't the only exception, and partly because we don't know it's black. Remember, in a black and white photo, light colors will often appear white, and dark colors will often appear black. It's not as though no other colors were worn before color photography was available! People wore a variety of colors, both light and dark, even during eras when one or the other was more popular.

In this historical clothing dating guide, you'll see if you scroll down all the way to the bottom of the 1900-1910 page, a group of women riding bicycles, all in light colored blouses except one. However, the one in a darker color is clearly not wearing a black shirt, as it's obviously a lighter color than her skirt. So she is not wearing black, which means she's probably not in mourning, but she is wearing a dark color, proving that not all women wore light colors all the time during this time period.

Consider also the impracticality of always wearing light colored shirts for a span of about ten years and trying to keep them clean in an era before the convenience of washing machines. In the same way that men of different stations in life would wear "blue collars" vs "white collars" depending on the type of work they did, so too would it have been impractical for certain women to always wear white or pastels. Even women of luxury probably still wore darker colors for certain activities, at least. In the cycling photo from the clothing dating guide (which I don't want to paste here because the copyright ownership is unclear), it seems that riding a bike would be one activity where it might be smart to wear a darker color.

Above (top left) is an image of my great grandmother, Emma Sarah Fallows (1888-1954). This was probably taken around 1900-1910, for a number of reasons. Her hair style, a typical Gibson girl style, was popular at this time, as was the bell shaped skirt she's wearing. It's a candid taken at her parent's home which was built in 1900, so it can't be before that year. The fact that it's a candid means it was probably taken with a Kodak Brownie, the first "snapshot" camera available/affordable to the general public, first released in 1900. But her clothing does not yet have the straighter waistlines that became popular for women in the 1910s, so it's probably pre-1910. As you can see, despite lighter blouses being more popular during this decade, she is wearing a darker shirt. Below are a couple more photos probably from the same day (she is wearing a cardigan but you can tell it's the same shirt, or in any case, it's a dark shirt). Note how her mother (Mary Ann Rorer) is wearing a lighter shirt. Same day/era, one woman wearing a light shirt, the other a dark shirt. There were no deaths in the family around this time, so no reason Emma would be wearing black for mourning, and even if she were, her mother likely would be as well. So don't get tripped up by thinking a popular trend meant no one wore anything else! That's not how fashion has ever really worked.