Ken, What I was envisioning in scenario 2 was the low but finite probability that even in an old clade some members would not have undergone much change in STR pattern from the clade founder, so that if one examined many members of the clade he would be expected to find a few who still resemble the clade founder and thus each other. Then, if only these few are analyzed for tmrca, and those who don't match are excluded, it seems that an erroneous value for tmrca would be obtained. In the case of the Z-139+ clade, which includes you, me, and AS2, maybe your argument is that the clade is too old and AS2 is too numerous to represent the founder's STR pattern or something close to it? Quoting Kenneth Nordtvedt <knordtvedt@bresnan.net>: I don't understand your scenario 2)? You have a bunch of haplotypes of a clade. They have accumulated STR repeat differences. Their variance is a measure of time back to the MRCA. That's it. When the ancestral line of the MRCA split from other branches of I1 is a different issue --- different node, as measured by interclade variances. -----Original Message----- From: Obed W Odom Sent: Wednesday, December 28, 2011 11:02 AM To: y-dna-haplogroup-i@rootsweb.com Subject: Re: [yDNAhgI] AS2 - Z138+ Ken, How can one distinguish between these 2 possibilities for AS2: 1) AS2 is a relatively young clade, as judged by its apparently recent tmrca or 2) AS2 is a part of a much older clade; namely, that part whose STR pattern has changed little (through the luck of the draw) from the pattern of the ancient common ancestor, while the STR pattern of the majority of this clade (again by purely random means) has changed enough that it no longer shows a close resemblance to the STR pattern of the common ancestor? It seems to me that in either case AS2 would *appear* to have a recent tmrca: in the first case because it actually does have a recent tmrca, and in the second case because it has been *selected* from the ancient clade precisely on the basis of a shared STR pattern, and the other members of the ancient clade have thus been discarded from the tmrca calculation. I'm having trouble seeing how these possibilities can be distinguished.

My haplotype and your haplotype are not members of clades as such; they are just part of the background. (That's not exactly true in my own case; I believe I have found two English haplotypes with common ancestry to my male line living 1300 years ago). Out of the many, many lines, there will be the whole spectrum of those closely matching the MRCA haplotype to those very different. There will be a distribution of differences from the founder which has a peak somewhere in the middle. If one picks subsets which somehow exclude the outliers or exclude the super central haplotypes, then yes, you will produce artificial tmrca estimates which are too low or too high, respectively. All I can do with my own haplotype right now is make tmrca estimates for the node between my line and the AS2 line, or between my line and your line, etc. AS2 is actually quite young. We don't know yet what makes up the rest of the Z138+ Z139+ population --- it could be just generics like you and me, or other clades identified on basis of their STRs could be included??? Ken -----Original Message----- From: owodom@mail.utexas.edu Sent: Wednesday, December 28, 2011 5:18 PM To: y-dna-haplogroup-i@rootsweb.com Subject: Re: [yDNAhgI] AS2 - Z138+ Ken, What I was envisioning in scenario 2 was the low but finite probability that even in an old clade some members would not have undergone much change in STR pattern from the clade founder, so that if one examined many members of the clade he would be expected to find a few who still resemble the clade founder and thus each other. Then, if only these few are analyzed for tmrca, and those who don't match are excluded, it seems that an erroneous value for tmrca would be obtained. In the case of the Z-139+ clade, which includes you, me, and AS2, maybe your argument is that the clade is too old and AS2 is too numerous to represent the founder's STR pattern or something close to it? Quoting Kenneth Nordtvedt <knordtvedt@bresnan.net>: I don't understand your scenario 2)? You have a bunch of haplotypes of a clade. They have accumulated STR repeat differences. Their variance is a measure of time back to the MRCA. That's it. When the ancestral line of the MRCA split from other branches of I1 is a different issue --- different node, as measured by interclade variances. -----Original Message----- From: Obed W Odom Sent: Wednesday, December 28, 2011 11:02 AM To: y-dna-haplogroup-i@rootsweb.com Subject: Re: [yDNAhgI] AS2 - Z138+ Ken, How can one distinguish between these 2 possibilities for AS2: 1) AS2 is a relatively young clade, as judged by its apparently recent tmrca or 2) AS2 is a part of a much older clade; namely, that part whose STR pattern has changed little (through the luck of the draw) from the pattern of the ancient common ancestor, while the STR pattern of the majority of this clade (again by purely random means) has changed enough that it no longer shows a close resemblance to the STR pattern of the common ancestor? It seems to me that in either case AS2 would *appear* to have a recent tmrca: in the first case because it actually does have a recent tmrca, and in the second case because it has been *selected* from the ancient clade precisely on the basis of a shared STR pattern, and the other members of the ancient clade have thus been discarded from the tmrca calculation. I'm having trouble seeing how these possibilities can be distinguished. ------------------------------- To unsubscribe from the list, please send an email to Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' without the quotes in the subject and the body of the message