How Accurate are the Programs that Find Names in the Family Tree? Part One

Disclaimer: I am going to write about some genealogy programs without mentioning the names of the programs. The reason for this is simple. I consider these programs as a category of programs and I do not wish to imply that any one of them is either better or worse than the others. This is not a review of the programs themselves, but of the basic concept that has been used to develop them all.

These programs all search the FamilySearch.org Family Tree for "temple opportunities." Essentially, they are searching for a specific category of people within the Family Tree; those that have "green temple icons." Assuming that FamilySearch has used some criteria for marking individuals in the Family Tree as available for temple ordinances, it is really a trivial programming issue to search through a database such as the Family Tree for such markers. For example, let's suppose I want to search through the following list of numbers:

1, 2, 3, 4, 5, 7, 8, 9, 10

If I consider the sequence, I can readily determine that there is a missing number in the sequence. But for a computer programmer, the problem would more likely be resolvable by creating a field for numbers and then searching the fields.

Enter your number here _____

In this case, even if the numbers were not sequential, all I would have to do is search for an empty field.

However, unless the computer program was extremely sophisticated, the program would not consider the "reason" for the empty field. In the case of the searches of the Family Tree, the opposite of looking for an empty field is looking for a field that has a specific marker. For example, if I want to know how many males and how many females there are in the database, I can create a required field for marking the sex of the person. It is then rather simple to write a program that counts the number of males and females. The program could also mark how many of the fields were empty and have another simple process, and if/then condition that produces an error message telling the programmer that the field is empty.

OK, so that may not sound trivial but it is. So, if someone tells me that they have written a program that searches a database and tells me if certain information contained in a field is missing, I have to believe that this is a rather simple operation and the results will be trivial. These types of programs are the same as those that notify you when you receive an email message or tell you to answer your cell phone when someone calls. They answer yes/no questions.

Is the fact that an entry in the Family Tree is missing temple ordinance information a simple operation such as the ones I just described? Not at all. In the case of missing temple ordinance information, i.e. a green temple icon, there is a whole hierarchy of considerations such as some of the following:

• Is the entry in the Family Tree correctly identified?
• Is the entry a duplicate of some other entry?
• Is the entry complete enough to allow temple ordinances to be performed?
• Does the person identified in the Family Tree qualify for temple ordinances?

But ultimately, the whole issue boils down to whether or not the person who is viewing and using the Family Tree is one of the people who can perform the ordinances pursuant to a complex system of qualifying people to do temple ordinances in the first place. 

For example, let's suppose that there is a rule that only people with a certain type of relationship to the individual in the Family Tree can perform the ordinances on that person's behalf. This is actually the case when you get the following type of warning:

So, the rules involved in determining whether or not a person marked with a green temple icon is actually available for ordinance work is more complicated than a simple yes/no program for searching a database. 

But as I pointed out above, one of the main issues with determining whether or not ordinances can be performed is the issue of identification. If I add a random name to the Family Tree, the program will immediately mark that name as available for temple ordinances as long as there are entries in the required fields. For example, if I make up the name of a child in a family that did not exist and provide made-up birth and death information, the program will not detect the fraud. But there will be a green temple icon generated for the bogus entry. It doesn't really matter whether I do this intentionally or unknowingly or even accidentally, the computer program does not detect the misinformation or fraud. So how do we know if the green temple icons are accurate indications for the need for temple work? The simple answer is that we cannot rely on the programs.

But wait, the issue is even more complex that a simple addition of wrong information into the family tree. The Family Tree is lineage linked. This means that the Family Tree has information that should be reflecting my ancestry. But what happens if one of the lineal entries is inaccurate? Then all the information following the incorrect entry is also inaccurate. So for example, if my great-grandfather is incorrectly identified in the Family Tree then the person shown as my great-great-grandfather is also wrong, even if the connection between the two is correct. One mistaken identity entry negates the accuracy of all of the entries dependent on that entry. 

One inaccurate or false entry in a pedigree line negates the accuracy of all of the dependent entries. I could repeat that over and over again in fifty different ways and it would still be an expression of the basic assumption of the Family Tree. So, when the Family Tree has a green temple icon, in effect, it is a statement that all the entries in the Family Tree connecting you to this ancestor or relative are "correct." The problem, of course, is that in many cases, this information is not correct. Over a hundred years of accumulated errors as well as accurate information is contained in the Family Tree. 

What is the effect of the unreliability of the Family Tree? Some of the results are trivial and some are serious issues. Here are a few of what happens when we unknowingly rely on the accuracy of the Family Tree:

• Ordinances are duplicated
• Ordinances are performed for people to whom we are not related
• Ordinances are performed for people who did not exist

As I have pointed out many times previously, in any system such as the Family Tree, the designers or developers, as engineers and computer technicians work within a system that allows for a degree of error or inaccuracy usually referred to as the margin of error defined as an amount (usually small) that is allowed for in case of miscalculation or change of circumstances. Since it is a given that the Family Tree has errors, the fact that some of the green temple icons indicate the classes of people who are shown in the short list above, falls within this "acceptable" margin of error. 

In the case of the Family Tree, the programming has become more and more sophisticated over time. Many of the duplications, errors and issues with relationship have been addressed by the programmers. Compared to just a few years ago, the Family Tree is markedly more accurate than it was at the time the transfer was made from new.FamilySearch.org to the Family Tree. But have we come far enough to rely on simple search programs to provide us with "temple opportunities?"

Now we come to the issue of the programs that search the Family Tree for green temple icons. But I will have to defer further writing until another blog post. 

Stay tuned.