Jump to content

Talk:Pedigree collapse

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Alfonso XIII

[edit]

Rmhermen asked me the source for my change of Alf's gggparent's count from eight to ten. I first noticed that the Snopes article linked to from the Pedigree collapse page said ten, not eight, gggparents. I then found this page: [1] which showed a full pedigree. It shows his father had 4 ggparents and his mother had 6 ggparents. (Of course, his father, Alfonso XII, had no more than 8 gggparents by this reckoning.) -R. S. Shaw 22:01, 2004 Nov 29 (UTC)

Regarding "Some geneticists believe that everybody on earth is at least 50th cousin to everybody else."

I User:jamesdowallen challenge this assertion. Most pairs of people are much closer than this; some pairs are much further! First, note that cousins grow (in simplistic model) by a factor of 4 per generation (not factor of 2 as for ancestors), so one would have 4 billion 15th cousins!" This ignores pedigree collapse, but a good guess would be that any two Europeans are closer than 20th cousins. As for "everybody on earth", what about the relation between, say, an Australian aborigine and an Amerindian deep in Amazon jungle? Is there any reason to suppose all such people have post-Columbian Eurasian ancestors? Is there any guarantee of trans-Pacific contact after early migrations? "50th cousin" implies the two isolated aborigines would have a common ancestor within the last 2000 years (approx). Is there any reason to believe this?

Please read up on this at Most recent common ancestor and its linked pages. Rmhermen 13:22, 11 January 2006 (UTC)[reply]
Also, that's why the sentence says some geneticists believe ... It's not presented as a hard fact, it's an expression of a hypothesis/theory held by a reasonable number of people in the field. If it said Everybody on earth is at least 50th cousin to everybody else then your challenge would be very appropriate. - DavidWBrooks 14:20, 11 January 2006 (UTC)[reply]
No credible geneticist would claim this. Fifty generations is not very long ago, maybe 1500 years. If everyone's ancestors wandered around so widely, and interbred so freely, that everyone in China and Congo and Paraguay is related to everybody else, then people would not look different and there would not be so many different languages. If you want to claim that "some geneticist believe this", then cite a geneticist who actually does, not some bogus website.Lathamibird (talk) 19:44, 23 October 2016 (UTC)[reply]
Wow - a response to a comment written 10 1/2 years ago!
Did you read the bogus website? It cites a number of excellent sources, although not clickable (i.e., printed books). However, you have a point. I am unable to Google-search any first level source for the claim (although a zillion places repeat it) - which is more of an indication than it would have been 10 years ago.
Further, I found this argument (http://walkitout.livejournal.com/700663.html) that pokes holes in one of the originators, or perhaps *the* originator, of the 50th-cousin claim. - DavidWBrooks (talk) 21:41, 23 October 2016 (UTC)[reply]
Renowned geneticist Richard Dawkins notes, in the Ancestor's Tale, how surprisingly few generations separate us from cousins across the planet. Published sources by experts using estimates of the rate of occasional breeding across vast distances even for isolated populations, along with statistics and simulations, support the idea that the number is indeed surprisingly small, e.g. 76 generations. See details at most recent common ancestor, and good explanations and sources at https://humphrysfamilytree.com/ca.math.html, and my recent edit here to more recent studies.★NealMcB★ (talk) 02:31, 14 September 2020 (UTC)[reply]

Which was the inbred Habsburg whose "ancestry overlap" was higher than if his parents had been full siblings? —Preceding unsigned comment added by 83.104.132.41 (talk) 11:27, 17 June 2009 (UTC)[reply]

Simple example of a collapse

[edit]

What do people think about adding a simple example to the article? such as:

Mary, John, Bob and Carol each have two parents, four grandparents, eight great-grandparents, which would bring the total of great-grandparents bewteen the four to 32, except for the fact that Mary, John, Bob and Carol are all brothers and sisters, there for 32 collapses to only eight great-grandparents total due to common ancestry.▪◦▪≡ЅiREX≡Talk 23:57, 8 October 2007 (UTC)[reply]

The example suggested (using four siblings) has no pedigree collapse. A person has (normally) eight great-grandparents, and any siblings (be they Mary, John, Bob, Carol or whoever) have the exact same parentage and therefore the exact same great-grandparents - eight, no more and no less, and therefore no pedigree collapse. It would require M/J/B/C's parents to be first cousins (sharing one pair of grandparents in common) for a pedigree collapse to exist. --Pete Hobbs (talk) 12:43, 5 July 2011 (UTC)[reply]

German Article that links to this Pedigree Collapse

[edit]

I User:la loi du récit suggest someone translate most of the German article that translates to this article. I believe a great deal is lost in the translation between these two articles, and I don't see how they easily correspond. If someone can suggest how I might find this translation in work in a different article, then link me to that article please. laloidurécit 31 October 2008 16:46 Eastern

There is some intriguing material in the german article - in particular suggesting Augustus II the Strong had only 12% of the maximum number of ancestors going back 12 generations (499 vs 4096). But I don't see a source of their cool table. ★NealMcB★ (talk) 02:36, 14 September 2020 (UTC)[reply]

Elizabeth II and Charles II

[edit]

Like most people, the family tree of Elizabeth II to five generations has 62 different people in the 62 different positions. The family tree of Charles II had only 32 different persons in the 62 positions. Going back two more generations, he had only 82 different people in 254 positions.

Going back five generations, you have 32 positions. Six generations is 64 (not 62) positions, while eight generations is 256 positions (not 254). I've corrected those numbers, but are the "32 different persons" and "82 different persons" part correct or also mistakes? XinaNicole (talk) 08:54, 1 February 2010 (UTC)[reply]

You did the math wrong. It is not 2 to the 6th=64 like you think. It is 2+4+8+16+32. Look at this ancestry box to the sixth generation to see. There is only 62 slots plus the person.--KAVEBEAR (talk) 06:19, 12 May 2010 (UTC)[reply]
Firstly, this talk-section should have been titled "Elizabeth II and Charles II" for clarity, as it discusses two rulers (not just Charles II), and I have therefore amended the title accordingly. Secondly, the original description for Elizabeth II (pre-Feb.2010, as quoted above) was correct before alteration to a currently wrong and confusing "The family tree of Elizabeth II to six generations has 64 different people in the 64 different positions." The chart on a Wikipedia page titled Ancestry_of_Elizabeth_II shows quite clearly that she has 62 different ancestors in the FIVE immediately preceding generations of her family - exactly as per the original description. As for Charles II, the same 2010 revision to "The family tree of Charles II had only 32 different persons in these 64 positions; going back two more generations, he had only 82 different people in the 256 positions" is equally incorrect, as the total number of possible positions can only be 254. Evidence of the 82 different people would be nice, if someone could find/add a link. Meanwhile I shall restore the originally quoted sentences immediately. --Pete Hobbs (talk) 13:58, 5 July 2011 (UTC)[reply]

Implex

[edit]

It is unclear what this word means in English but it does not appear to adequately cover the entire topic here (called pedigree collapse). [2] say an implex is "a person who appears several times in the ancestor tree" while the implex factor is (total possible ancestor count - actual)/ possible. For the writers of MaqPar[3], it is something different and they have three different kinds and acknowledge using the word differently from "the old-guard Brazilian genealogists". [4] uses "ancestors' implex" for what is apparently implex coefficient elsewhere. [5] seems to use the term to indicate a single location of collapse in a tree. (19:46, August 6, 2010 Rmhermen)

See wikt:implex. The implex is the ratio of virtual to real ancestors, i. e., describes the same phenomenon, only expressed as a number, while "pedigree collapse" is a metaphor simply referring to the phenomenon as such. That said, I think the definition is incomplete. "Implex" originally means something like "entangled", and the number is the "implex coefficient", strictly speaking, though it may be referred to as "the implex" for succinctness. --Florian Blaschke (talk) 01:05, 26 July 2013 (UTC)[reply]
I note above multiple people using differing definitions. Rmhermen (talk) 01:43, 26 July 2013 (UTC)[reply]
All the links are now dead, though, except the Swiss one, which is not very informative and does not contradict my statement (and the definition in Wiktionary) anyway. The others may not have been reliable (professional/academic) sources in the first place. I'd recommend trying to find a reliable source; an introduction to genealogy would be your best bet.
That said, I've found a glossary explaining the term in my sense here, on the website of a professional genealogist, which may count for something. --Florian Blaschke (talk) 13:41, 29 July 2013 (UTC)[reply]
I am not certain that this word has a single well-defined meaning for everyone. One of the links (MaqPar) can be seen at [6] now.
I have removed it from the introduction, since I can't find any good references that support saying it is "also known as implex." If it is used by anybody in this fashion, it doesn't seem (without better references) to be used enough to be worthy of the introductory sentence, at this time anyway. - DavidWBrooks (talk) 15:49, 29 July 2013 (UTC)[reply]
MaqPar does not contradict wikt:implex, but again, this is strictly speaking the implex coefficient or implex factor, not the phenomenon of implex itself.
The Genealogical Dictionary, which supports my position, was published in book form in 1998, as its author, Felix Gundacker, has kindly informed me. Therefore it is citeable and I will re-introduce mention of the term implex, especially seeing that Implex still redirects here. --Florian Blaschke (talk) 10:22, 8 August 2013 (UTC)[reply]

Revision of "How it works" section

[edit]

I revised this section, maintaining the references and the general and specific example of royalty Alfonso XII. I removed the more arcane mathematical observations and replaced them with straightforward examples.

In a school paper or a conference paper, the idea is to impress knowledgable, educated readers. In an encyclopedia, and especially in Wikipedia, the idea is make text easily understood by the likely readers. In this case, that might be almost anybody investigating genealogy. So use of phrases such as "directed acyclic graph" and "binary tree" and "High Middle Ages" is counter-indicated, since probably only 5% of readers would know what those all are. 98.210.208.107 (talk) 01:19, 1 March 2011 (UTC)[reply]

Nicely done. Clear and, so far as I can tell, quite accurate. - DavidWBrooks (talk) 01:31, 1 March 2011 (UTC)[reply]
I disagree (mostly). Conference papers are not written to impress, they are written to spread information. School papers are (in a proper school) an excercise in doing so. Dividing written material into two categories, one to impress its readers, the other to be easily understood and then assigning wikipedia articles to this second category is specious. Furthermore, in an online encyclopedia such as wikipedia, the reader can simply click to open an article on any topic that is unfamiliar, such as those mentioned above. So concepts that may be difficult to the casual reader still have a place in the article. If a version without any such concepts is still deemed useful, then it can be placed on simple.wikipedia.org. It is however a good point that such concepts should preferably be used in a manner that still allows a reader to understand the article also when those concepts are unfamiliar. Reinstating directed acyclic graph accordingly - after all, readers unfamiliar with graph theory might find this real world application of it interesting. Lklundin (talk) 18:45, 1 August 2011 (UTC)[reply]

Measure of Pedigree Collapse

[edit]

Thanks LKlundin for maintaining the "binary graph", "directed acyclic graph" and "High Middle Ages terms".

This article would benefit greatly from a measure of how much pedigree collapse typically takes place after n generations. Presumably, for most people there is 0 collapse thru generations 3 or 4 but by generation 10 on average, how much collapse has taken place? By generation 20? Skates61 (talk) 00:29, 30 December 2011 (UTC)[reply]

:"If a child and parent were to procreate, their offspring would have four grandparents, although one of these would also be a parent and therefore introduce no additional genes – thus procreation between parents and children would result in less pedigree collapse than procreation between full siblings." How can the fact that no new genes are introduced result in less pedigree collapse. That's what it says. CapnZapp (talk) 16:01, 1 March 2016 (UTC)[reply]

From my point of view the sentence in this Wikipedia article "If a child and parent to procreate ..." is complete nonsense. --Hermannh (talk) 17:17, 6 April 2024 (UTC)[reply]