Wednesday, March 6, 2013

Sorting Out Primary and Secondary Information

One of the quality measures to be considered in the Genealogical Proof Standard is the issue of Information provided by a source record - is it primary or secondary information?  The definition of these two categories is (from the BCG article here):

*  Primary information:  Recorded by a knowledgeable eyewitness or participant in that event, or by an official whose duties require him or her to make an accurate record of the event when it occurs.

*  Secondary information:  Supplied by someone who was not at the event and may include errors caused by memory loss or influenced by other parties who may have a bias or be under emotional stress.

I have been struggling with this issue for awhile - and thought that I would throw out an example and talk my way through it, then ask my readers for their comments and opinions:

I'm going to use the birth record of my second great-grandfather, Isaac Seaver (1823-1901) as the example because I have at least four Source records for his birth that can provide useful discussion.

1)  This is the Westminster, Massachusetts town clerk's record book that records the birth of Isaac Seaver on 16 October 1823:

I don't know who the informant for this record of Isaac's birth was - my best guess is that his father, Benjamin Seaver, visited the town clerk soon after the event and said something like "my wife had a son and we named him Isaac, born 16 October 1823."  The informant was very likely knowledgeable about the birth, and may have witnessed it or become aware of it minutes or hours after the event.

My conclusions:

*  This record is an ORIGINAL source.
*  This record is PRIMARY information (provided by someone with first-hand knowledge of the event).
*  This record is DIRECT EVIDENCE of the event.

2)  In 1908, the Westminster, Massachusetts Vital Records to the Year 1849 book was published by F.P. Rice (commonly called the Tan Books).  Here is the page recording the birth of Isaac Seaver:

The person who extracted the information from the town records in 1908 is the informant for this record - someone who was not present or had first-hand knowledge of the birth event.

My conclusions:

*  This record is a DERIVATIVE source.
*  This record is SECONDARY information (provided by a person who was not a participant or eyewitness, and who extracted it from the town records and published it).
*  This record is DIRECT EVIDENCE of the event.

3)  In 1990, I wrote to the Westminster, Massachusetts Town Clerk's office and obtained a birth certificate for Isaac Seaver:

The town clerk in 1990 probably consulted the town record book (she listed the year, volume and page number in the lower left-hand corner), filled out the birth certificate and certified her action with a seal and signature.  The town clerk is the informant for this record.

My conclusions:

*  This record is a Record Copy of an ORIGINAL DERIVATIVE source (since it was created 167 years after the town clerk's book)
*  This record is SECONDARY information (provided by a person who was not a participant or eyewitness, and extracted it from the town record and certified its accuracy).
*  This record is DIRECT EVIDENCE of the event.

4)  My quandary has been:  The Town Record (item 1) is the only information that was provided by an eyewitness or participant in the event, and the 1823 town clerk was an official required to make an accurate record of the event.  But the other two sources used the town record book (my assumption, but I think it's realistic), but were not an eyewitness, participant or official who recorded it.  Should the information in items 2) and 3) above be considered PRIMARY information or SECONDARY information?  

Should item 3) be considered PRIMARY information because the informant (the 1990 town clerk) was an official required to record accurate information?  I don't think so.

My conclusion is that items 2) and 3) are SECONDARY information, since the informant for the record is not an eyewitness, participant or official recorder at the time of the event.

What do you, my knowledgeable and faithful readers think?  Have I scoped this out correctly?  Am I worrying too much?  Is this obvious?  Your comments are desired and appreciated!!

The URL for this post is:

Copyright (c) 2013, Randall J. Seaver

UPDATED:  8 March, to correct the third document's Source label


Unknown said...


Your logic is pretty sound on all counts except one (maybe two). Because the informant for the entry in the town record book isn't specifically identified, you can't be sure that it's primary information.

I notice that the pages are divided into families, then events are listed chronologically for each. That doesn't even indicate that the entries were made at or near the time of the event - just that they were made some time before the next event.

In my opinion, the first source should be considered original (although the consistency of the handwriting over a period of over 40 years leads me to believe that this book could have been copied and divided into families). I would classify the information not as primary or secondary, but as UNKNOWN. Dad might have given the information, but he might not. In any case, it is direct evidence of the event.

I'm curious to hear other people's opinions on this, as well as on the fact that the 2nd and 3rd source appear to simply be derived from the 1st source - how would you handle that in your analysis?


JG in MD said...

I'm a simple person. I'm not a teacher and I tend not to label data unless I need to. If all these sources agree, I'm afraid I'd never give a thought to the classifications. If they differ significantly, analyzing begins and so does classifying.

Sonja Hunter said...

I agree that the first source seems to be the best one. However, I would classify the third source as a transcription and not as a true copy. Maybe I'm being nitpicky here, but I would consider a photograph or a photocopy as an actual copy. If you didn't have the first two sources you would never know if there was a transcription error. Lorine at Olive Tree Genealogy ran into exactly this problem when a record she requested from the state of Michigan was transcribed incorrectly.

Bob McAllister said...

At least part of the difficulty you mention comes from the fact that the first two dimensions are not independent.

Once you decide that the source is derivative, then the information can not be primary. If an eye-witness to the event made a copy of the original source, it would itself be an original source with the same provenance as the first (save for the lapse of time).

It is sleight of hand to claim that we have analysed each document on three distinct criteria when two of them are inextricably intertwined.

Unknown said...

Randy, I'd bet your first record is actually a copy as Jenny suggested. In that case they are all secondary.

Sue Adams said...

I think people can get too hung up on the classification. It is usefull in encouraging us to think through the processes of just how each document was created and copied. The observations that the first record may have been a copy because of the handwriting consistency and that the informant could have been a witness to the birth are part of the analysis of record's reliablity and I think more informative that a 3 box classification.

The classification is a tool, not an end product.

Susi's Quarter said...

Randy, I have encountered this many times since all this new classifications came about. My thoughts run like yours, confused but the bottom line is the data is the same. If a date or data was different then I think is when we should worry. To intense and takes away all the fun, sad to say.

Anonymous said...

Hi, Randy,
I agree that the first source you post does not seem to be a primary source as a book of registers would typically not be broken up into families.

I differ in my opinion from Sonja, however, in that I believe the third source to be an abstract, not a transcription. The information was not transcribed as it was in the record, but was abstracted to fill in the form. Either way it is not a copy of an original.

I also agree with the "unknown" designation for the first record because you do not know for sure whom the informant was. It could have been an aunt, neighbor, etc. (though not likely) that reported on behalf of the family but was not present at the event.

Fun stuff, this is!!

Terri Carlson said...

My niece is applying for membership in the Mayflower Society using my father’s ancestry as the basis. Her initial application was rejected because she submitted a Certified Copy of Birth Registration for my father. The argument was that “the registration [only] proves that the birth was reported”.
I subsequently obtained a Certified Copy of Live Birth for my father. The information on the CCLB reported EXACTLY the same information as the CCBR. The only differences were administrative (date, Registrar name, etc.).
The CCLB is considered acceptable as evidence for Mayflower Society membership even though it does not provide any additional vital information. By your definition, are they not both primary records?

Unknown said...

Terri, I think the reason the certified copy of certificate of live birth wins that race is because the doctor (or midwife or whoever was present) had to fill it out at or near the time of the birth.

The registration could have happened anytime after that, and the information in that case is no longer primary because the clerk was likely not present at the birth.

Geolover said...

I think Jenny Lanctot nailed it. The front cover of the book entitles it 'Births, Deaths and Intentions' and has a notation, "bought 1786."

However, earlier entries are decades earlier than 1786 in date, indicating that there probably was at least one other record. The index at the front has page numbers in no numerical sequence at all, which also suggests copying from other records at a lot of different dates.

One pattern I have seen in some records is recording of children born previously (elsewhere) when a family moves to a new place. In such an instance one of the parents can be assumed to be informant, but surrounding evidence must be gathered to corroborate when the family appeared in this place and united with a congregation.

This record book is part of a collection of microfilms of selected documents. The collection was not made as part of an effort to microfilm ~all~ pertinent documents. A check of the Massachusetts State Library website or query to the Town Clerk might indicate what records for this period actually exist.

Eric Jorgensen said...

I would classify the first record as either Secondary or Unknown. While the information may have been from the father (or mother), it wasn't specified, so we cannot know for certain (Unknown). But because the information was written into the record by someone separate from the event (the clerk), there could have been errors, misspellings, etc. (Secondary)

How often do we see Census Enumerators ("officials...require[d] make an accurate record of the event") misspell or introduce errors?

And I would concur with others that the third is an asbtract, and therefore derivative (not a copy of the original, as it was changed to fill a form). Even if judging the actual paper received in 1990 as it's own "original", it would certainly be Secondary as it's removed by a century and a half.

Yvette Hoitink said...

To my understanding, original or derivative tells you whether the source is in its first recorded form while primary or secondary tells you whether the contents were given by somebody who had firsthand knowledge of the event shortly afterwards. The distinction is used to assess different aspects of reliability. For example: in an abstract of a birth record you must look out for typos or information that was left out, but in a original death record you must be aware that the information about the birth of the person was not recent and probably not by a reliable informant.

In other words: transcribing or copying a record would not effect whether the information is primary or secondary, just whether the source is original or a derivative. Otherwise it would not make sense to have these different aspects.

In my opinion, all of these records contain primary information, but all of them are derivative as even the town book does not appear to be the first recorded form.

Michael Hait said...

Yvette makes the distinction correctly in regard to your second source. The author of the book is not the informant for the information--any more than the clerk is the informant for a recorded deed or the enumerator is the informant for a census record.

It appears that the second book is probably a modern transcription/abstraction of the first book.

However, the first book would also accurately be called a derivative source, because births were not recorded *in that volume* contemporary to the events. They were probably transcribed from an earlier register to put them in order. There may be an earlier underlying record set to be consulted.

Likewise, as Jenny stated in the first comment, none of the records identify the informant directly, so whether the information is primary or secondary is UNKNOWN. You may be able to discern who the informant likely was, based on both internal and external factors, but there will never be certainty to this point.

Lastly, while making the distinction between original/derivative sources and primary/secondary information can be helpful, it can also be a bit arbitrary. The point is not so much whether a source is one or the other, but what its being one or other tells about its quality as a source of information. There are many other factors to consider in the process of analysis, some of which may be far more significant.