Alan, When I first drew attention to the fact that the concept of “negative proof†does not exist but is simply an estimation of probability, I certainly was not suggesting that we should start to calculate that probability either using Bayesian theory or otherwise. Indeed, I can see this this producing a row of blank faces, and putting people off for life :-). For me, Legacy’s validity assessment is good enough, with a maximum value of 2.
The point is that it should be accepted that there is always some doubt when a relationship is derived by a process of eliminating all other options, and such a relationship can never be set in stone. Ron Ferguson, http://www.fergys.co.uk/ From: Alan Pereira Sent: Wednesday, August 15, 2012 10:31 AM To: LegacyUserGroup@LegacyUsers.com Subject: RE: [LegacyUG] Proof of relationship Brian, I have resolved similar issues by finding the descendants of All 3 brothers. You may find a namesake of your 6x GGfather with the other families thus eliminating him from that line (negative proof again). I take kb's view that assigning a Bayesian probability may be a pragmatic approach to identifying the value of the justification in making this assumption. I am not sure that I want to go into the math to provide a quantitive value or that a later descendant would have any hope of understanding it! Alan From: Brian Woolvett [mailto:woolv...@one-name.org] Sent: 15 August 2012 02:23 To: LegacyUserGroup@LegacyUsers.com Subject: RE: [LegacyUG] Proof of relationship I find this thread very interesting and informative, thanks to Kb and Ian for their comments I am in a similar situation of my 6x Great Grandfather being the probable son of 1 of 3 brothers, I will probably never be able to prove who the father is but the idea of bypassing a generation with a probable is giving me a way out. Taking the argument a bit further how many of us can say without any shadow of doubt that our father who is on our birth certificates is true!! Thanks again Brian Woolvett South Australia www.woolvett.com From: Ian GARDENER [mailto:ijg3...@gmail.com] Sent: Wednesday, 15 August 2012 8:55 AM To: LegacyUserGroup@LegacyUsers.com Subject: RE: [LegacyUG] Proof of relationship Hi kb, Thanks for the tip on Bayesian probability. I must confess that the math is way beyond me these days (maybe alwaysJ) but the underlying broad principle is definitely something I can run with for some of my research. I agree that Legacy seems quite weak in the research log, notes department and it something I hope is improved in version 8. In the meantime I just make my notes in word then create a pdf file which I attach to legacy sourcing. I don’t use native word format because the file structure is everchanging. I like your idea of “demonstrating†rather than necessarily proving relationships as I too am really laying groundwork for future family members I hope. For that reason I also make & update hundreds of pdf reports as the raw data may also become useless in the end. Thanks again. You’ve given me a kick toward detailing my reasoning in more detail. Ian GARDENER Australia www.gardener.org.au From: britton...@comcast.net [mailto:britton...@comcast.net] Sent: Wednesday, 15 August 2012 4:37 AM To: LegacyUserGroup@LegacyUsers.com Subject: Re: [LegacyUG] Proof of relationship Alan, I also just document reasoning in research notes. Legacy is excellent for organising classical genealogy material, but it's limited or poor as a research tool or record. It has little support for anything less than fact, and that bites not only subsequent descendants, but also the present genealogy researcher. I've sympathy for Ron's position regarding "negative proof", but some term is needed. On balance, I see the trade off, between violating prior use in information theory versus important advantage to clarity in your usage, as decidedly on your side. I'm leaving history and genealogy of my line(s) to the future, but also everything I can find which may be useful to subsequent or collateral researchers. Necessarily, that includes not just my logic behind judgements made, but also conjecture, unconnected fact, and rejected "fact" with reasoning. Generally, I support the position taken in the genealogy with detailed reasoning, followed by "Alternative possibilities" with discussion on pros and cons in less detail. Implicit in that is acceptance and warning that my adopted position may be wrong, but a "Negative Proof" textual label is a much stronger message that the adopted position is not only provisional but recognised by the author as in significant doubt. (Thanks Alan) If we confine ourselves to "classical fact", we don't just constrain ourselves to documents; we risk crippling our reasoning. Scientific research proceeds from hypothesis to proof, yea or nay. Hypothesis is, by definition, provisional and uncertain. If we don't keep clear awareness of "nay", we too easily seek only evidence to justify assumptions. Legacy seems to have few formal places for "provisional fact", and genealogies contributed to the web are typically stripped of research and other cautionary notes. I much recommend looking up "Bayesian" in Wikipedia. Crudely estimating probabilities and using them as in Bayesian Inference can be a help and a powerful discipline in research guidance. A quick check shows an average of 8 Dave XXXs alive in a parish at any time over 200 years. Not good odds that any record is of the one you want. It's a caution and focuses work on factors which improve probability. Unlike other logical systems, Bayesian can be reversible without circular reasoning. In retrospect, the odds on truth for an initial premise may be drastically higher when calculated in reverse from findings, and a traditional "fact" may turn up, permitting the discovery process to be discarded. How, in Legacy, do we keep track and record? Recently, I was contacted about a post I made in 2004, on a line I later dropped. My Legacy file had a few names and notes of interest to my contact, but resort to paper files and old disks from long dead computers produced a gold mine of lines of attack and disparate data for him, including emails from folks now dead. When I'm dead too, my Legacy file will still be on several computers, but my other records will be gone. Legacy needs a "research companion" with cross referencing from the standard program, but is it impossible to somehow formalise "negative proof" discussion within Legacy fields and facilities? kb From: "Alan Pereira" <alanpere...@tiscali.co.uk> To: LegacyUserGroup@LegacyUsers.com Sent: Tuesday, August 14, 2012 4:35:40 AM Subject: RE: [LegacyUG] Proof of relationship Thanks Charani & Ron I think I will just use the research notes when I need to clarify proof. Ron, I will adopt your probability approach when I have "negative proofs" as I agree entirely with you on that. Why I am doing this... I am trying to look at my Family Tree as if a descendant 2+ generations later picked up the tree and wondered why I chosen certain parents when there were other candidates. Also I am aware that variations on some lines in my tree exist on the web that I do not agree with (especially in ancestry). I want to document exactly why I made my choice, which for the most part rely on positive sources (baptismal and marriage records). Where I have the 'Negative proof' what I can do is prove it is Not the choice others have made in their trees. Taking on board Ron's comments this would increase the probability of my choice but not prove it. I gave up long ago in trying to correspond with people who showed these erroneous links, as their response, if any, was to cite the other trees that greed with them. Alan -----Original Message----- From: Ron Ferguson [mailto:ronfergy....@tiscali.co.uk] Sent: 13 August 2012 19:20 To: LegacyUserGroup@LegacyUsers.com Subject: Re: [LegacyUG] Proof of relationship 3rd time lucky!! Alan, I can see where you are coming from, but I have not found the need to do extensive work on this, with respect to pretty well all my family, who are readily identifiable from the sources available. There is one exception (my 5 G G grandfather) where I have been unable to find a source which links him to the person who I feel sure is his father. I have an extensive ToDo on him detailing all the sources I have checked, and when, and I still regularly make checks to see if anything has cropped up. The person I am sure is his father is there because there is nobody else I can see who could be, and other evidence indicates he is the most likely candidate anyhow. However, I do not accept the concept of "negative proof". This is often a concept used in ignorance by the media, often in the concept of a new medical treatment, asking for proof that there are no adverse effects. This can *never* be proven, only the probability range can be stated. Similarly one can never say that negative proof can be used to define a relationship - only the likelihood. In the case of my relative, it is possible, albeit unlikely given other evidence, that his parents were just passing through the town at the time he was born. DNA may resolve the question, but at the moment I have no contact who would serve to confirm the relationship. Ron Ferguson http://www.fergys.co.uk/ -----Original Message----- From: Ron Ferguson Sent: Monday, August 13, 2012 6:13 PM To: LegacyUserGroup@LegacyUsers.com Subject: Re: [LegacyUG] Proof of relationship Alan, I sent this earlier, before Charani replied, actually I sent it from an email address not registered at Legacy so it bounced! Anyhow, I decided to send it as my reply is not quite the same. Ron Ferguson http://www.fergys.co.uk/ -----Original Message----- From: Alan Pereira Sent: Monday, August 13, 2012 2:28 PM To: LegacyUserGroup@LegacyUsers.com Subject: [LegacyUG] Proof of relationship I am starting a research task in the todo list in providing proof(s) of relationship, which can be through a multitude of sources, the most difficult being the negative proof. My initial throughts were to create a document as a backup source detailing these proofs. Then again, why not just use the research notes. Just wondering what others do... Alan Pereira Legacy User Group guidelines: http://www.LegacyFamilyTree.com/Etiquette.asp Archived messages after Nov. 21 2009: http://www.mail-archive.com/legacyusergroup@legacyusers.com/ Archived messages from old mail server - before Nov. 21 2009: http://www.mail-archive.com/legacyusergroup@legacyfamilytree.com/ Online technical support: http://www.LegacyFamilyTree.com/Help.asp Follow Legacy on Facebook (http://www.facebook.com/LegacyFamilyTree) and on our blog (http://news.LegacyFamilyTree.com). To unsubscribe: http://www.LegacyFamilyTree.com/LegacyLists.asp
