193,000 living cousins

I think that the author of that article didn't consider the fact that a lot of those relationships "collapse" into the same person... The same thing happens with ancestors - it also applies to living cousins...

From what I've seen - I've got a lot of cousins who are related to me in three or four different ways... My neighbor across the road is my 2nd, 3rd and 4th cousin on different blood lines.

At this point - I've found literally hundreds of relationships like that on my tree. From DNA tests where I know the relationships, I've seen "convergence" where the test will say "3rd cousin" but the person is really a 4th, 5th, and 5th cousin.

If we work off the assumption that "we are all connected" then we are all cousins out there eventually - in which case the 193,000 is probably low. From a practical standpoint though, I've got to believe that what we can prove from a paper trail and meaningful DNA testing would tend to make that number much lower since some of the relationship end up being the same person over and over again.

So that 193K is both too low (many "cousins" not yet identified) & too high ("pedigree collapse") at the same time?

There needs to be a better way to conceptualize these things besides counting. Some sort of network of webs with gates that link population groups?

It does seem like I managed to answer the question in mutually exclusive terms. I left out some of what I was thinking so I'll throw out some of what was going through my head when I answered just to be more clear. I'm not too good explaining things sometimes - I leave stuff laying around in my head that might not be obvious to anybody else listening to my ramble.

From the article it seems like they arbitrarily drew the line at sixth cousins which implies that a seventh cousin isn't a relative at all in their methodology...

Given that DNA is not doled out proportionately - but randomly and not prorated according to any specific relationship - even siblings sometimes get the luck of the draw and seem to get different ends of the gene pool - that seems a stretch, since any given seventh cousin might share other cousin relationships - and more DNA - then somebody else inside that sixth cousin circle...

(I keep trying to picture it as a bunch of Venn diagrams - but it gets messy in two dimensions really fast since you run out of places to put other people quickly...)

I guess they had to draw a line somewhere, but it seems like it was just for the sake of illustration. The idea that any random one out of 300 people you meet is possibly a relative seems like a limited view, but I think it was phrased that way just to be thought provoking instead of accurate.

Everybody you meet is a relative (unless they are not human). So - I guess I disagree with the premise of the article - hence my idea that the number is way too low.

The probability that a random stranger is a relative is always 100% - it is just a matte of "how related"... So - the odds aren't one in 300 - they are always one out of one...

Everybody you meet is most likely related to you in a multitude of ways - many different cousin relationships concurrently. Hence my idea that the relationships "collapse" for cousins just like ancestors... Pedigree collapse is ongoing - everybody's relationships collapse - we all become ancestors to descendants... So - the collapsing is ongoing...

I think it is more of a spectrum or a matter of degree then a clearly discrete event... Its not a matter of IF, but how much - the other random person is related to you.

If we could plot all of our relatives in three dimensions by known relationships - and then somehow color code the DNA markers for each person - that would probably give a more meaningful picture of "how similar" or "how different" we are to one another.

It would be a real bear to program software to do that, but it would probably be a more meaningful and practical definition of "related"...

If we took just one characteristic at a time - say eye color - and mapped that across all the "close" relationships (arbitrarily inside the sixth cousin boundary) I think we'd start to see patterns emerge. As other characteristics were added - it would get much more interesting over time...

Related in what way... Eye color - blood type - or whatever...

I think a lot of constructs like race and ethnicity are the result of trying to wrap our brains around the superficial patterns that exist in the observation of our experience. Its easier to grasp for stereotypes then try to do all the internal accounting to maintain the "related in what ways".

I think a lot of us do genealogy because we want to believe we are special and unique and exceptional and that our ancestors were as well.
One the other hand - I think we also want commonality - community - and consensus. We want to belong - we want a place - we want a family...

I think many of us want to believe that there are characteristics that transcend the superficial or obvious that contribute to the special, unique and exceptional, but we also want some kind of comfort in belonging to a bigger group.

So - finding commonality with a stranger - is satisfying in some fashion - and I think that the article was just aiming to fill some kind of "warm and fuzzy" feeling rather then aiming for any real accuracy of terms...

Thinking of 'conceptualization' of relatedness...

I started drawing 'line graphs' of relationships while helping find connections for AJ Jacobs Global Family Reunion ... in particular, for connections between myself and the 5 current Good Morning America (GMA) hosts.

Between any pair of living people, it is interesting to take the Geni-found path and draw a line based on the found linkages (parent= up, sibling&spouse=sideways, child=down. Spousal link can be noted special, as it 'breaks' or at least 'shortcuts' any blood/DNA linkage.)

In GMA example, I had a 'connection' to some that went "way up" and rather directly down, while others wandered mostly 'horizontal' to find a connection.

That approach could also be used to show different paths between the same two people (e.g. several 'closest' bloodline paths vs. spouse-traversal paths).

It would also work to show the different "removed" kind of cousin relationship; e.g.: a line that wiggles upward X steps, but only downward Y steps would show a "cousin X-Y removed" from you.

Dynamically, hovering or clicking on a 'node' on the line would reveal the name/dates of that person (e.g. mini-hover-card-bio).

=========================

That got me thinking about how one might better show the mutual connections between a group of people; that is, the pair-wise connections between (for example) each person and the other 5 in a group of 6.

One way could be to arrange the people as points on a circle. The 'pathways' would be shown as lines with varying thickness/density and color. The "shorter" the path, the 'heavier' the line. Perhaps think of the lines as being a collection of coiled springs. The fewer the connections, the "tighter" those springs will be pulling those two people together. The longer the path, the "looser" the connection between them.

I can even envision a dynamic graphic, where you 'highlight' one person (say, yourself) on the circle, and all the others "slide around" the circle based on the relative "tension" on the links between them, so that those nearest to 'you' on the circle would be the one with the shortest 'paths', while those more on the opposite side of the circle would have the longest 'paths' between you.

Highlighting a different person (as the 'focus' would then re-adjust (slide them around) based on the "tension" of the links between them and all the others. While the link between you and that new focus wouldn't change, the relative positioning on the circle might, because their paths would (most likely) be pairwise different "weights" than yours.

Of course, in a "static" view, one would just equally space the people around the circle and just use the "weight/tension" of the lines to reflect the relative length of the path between each pair. A click of an option could allow/disallow spousal traversals to see how that makes things different.

Another 'variation' of the circle view would be to put the one 'famous'/notable/focus' person in the center, with everyone else on 'spokes' of a wheel, with the thickness and/or length of the "spoke" based on the path between them.

The 'spokes' could "join" where people are in common. Thus, all the people connected via the father would 'come out' in one arm of a "Y" direction, while those via the mother by the other 'arm' of a "Y", and those from a sibling or child in still other directions.

The "spokes" would look more like a 'fission reaction' chain than a straight "wheel spoke", splitting off where each person's path diverged from those that started out in common.

Dan - yes, that's in line with computer networking theory, isn't it?

And the thicker the line, the thicker the line becomes (a magnet effect -- because we "know" who the "notable" is).