Not a Primary or Secondary Source, by Michael John Neill

Great-grandma’s death certificate is an excellent lecture example. Full of “unknowns,” it proves all of us have our research blanks. After one lecture, an attendee approached me and said I should not use the document as it was a secondary source. I indicated the document was not a primary or a secondary source, but provided both primary and secondary information. The attendee’s comment raised an issue that still confuses some genealogists: primary and secondary sources.

The death certificate used for the illustration is a grave disappointment. Ancestor Ida Sargent Trautvetter Miller is one of my brick walls. While the name of her father is listed on the document, all other parental information is listed as “unknown.” Like most records on my great-grandmother, it raises more questions than it answers. What is certain is that the document is not a primary or a secondary source.

Sources are not referred to as primary or secondary. It is the pieces of information they provide that are classified as either primary or secondary. Primary information is provided by someone who reasonably had firsthand knowledge of the event. Secondary information is provided by someone whose knowledge of data is not firsthand.

The death certificate is a source of many pieces of data; it states more than just one fact. Some facts took place a few days before the record was created. Some facts took place in another century.

To determine what information is primary and what information is secondary, it is helpful to know the name of the informant and his relationship to the deceased. Not all documents provide this information, but fortunately this one does. Only the informant’s name and address were listed: Carl Trautvetter, Loraine, Illinois. Carl’s relationship to Ida is not stated (he was her son). As I review the details provided on the certificate, I try and determine the likelihood that Carl had firsthand knowledge of the information given and how much time had passed between the event mentioned on the certificate and the recording of the certificate.

While the certificate does not provide the date the information was provided by Carl, it appears to have been no more than a week after her death and burial, and was probably recorded even sooner. Timing is key as memories can fade over time. Carl would have firsthand knowledge of his mother’s name, residence, marital status, occupation, color, and gender. The same is true for the name of her deceased husband, and her date and place of burial. He also probably knew at what point in time she stopped working as a “housewife” (probably the point at which her final illness truly incapacitated her). The other details are more problematic as they are things Carl likely did not know firsthand and either knew from what he had been told or conclusions he had reached.

Carl’s knowledge of his mother’s date of birth (and her age) was obtained because he learned it from somewhere. His knowledge of this date is not firsthand as he was not there. He also indicates Ida had been a housewife for forty years. I am always a little leery of time durations that could have been approximated (as many ending in a “0” are). This amount of time is very close to the number of years that had elapsed since Ida was first married. Again Carl’s knowledge of this information is secondary as well–people are not usually present at their parents’ first marriage.

The only parental information provided for Ida is the name of her father. Carl’s knowledge here is secondary as well. His memory of his grandfather probably was very distant and he only “knew” his grandfather was his grandfather because someone told him who is grandfather was.

The information Carl provided may very well have been completely accurate. Classifying information as primary or secondary is not the same as saying it is true or false. This classification must be done in concert with all other available documents, records, and sources. Primary information can be wrong and secondary information can be correct. This conclusion just cannot be made from one lone document.

On Ida’s death certificate there is one other person who provided information: the doctor. He is the informant for the date of death and the information about Ida’s medical condition and her cause of death. Hopefully the doctor is a primary source for this information. It was noted that the medical portion of the certificate (including the date and place of death) were filled out the day Ida died. It is difficult to get more contemporary than that.

Sources are not primary or secondary simply because they typically provide details that were known firsthand and those obtained elsewhere. Sources typically are classified as original or derivative. Ida’s death certificate is an original source. Her entry in the Illinois State Death Index, my transcription of her death certificate, and this article which discusses her death certificate are all derivative sources as they contain information extracted from the original source and in some cases contain interpretations as well.

Why all the bother?

It boils down to accuracy and how much credence we give to information provided by a source. When we take time to think about whether bits and pieces of information are primary or secondary, it gets us thinking about the potential accuracy of the information–who provided the information, and the circumstances under which it was provided. And such an analysis can only strengthen our research conclusions.

Those who want to see a copy of Ida’s death certificate can click here.

Michael John Neill is the Course I Coordinator at the Genealogical Institute of Mid America (GIMA) held annually in Springfield, Illinois, and is also on the faculty of Carl Sandburg College in Galesburg, Illinois. Michael is currently a member of the board of the Federation of Genealogical Societies (FGS) He conducts seminars and lectures nationally on a wide variety of genealogical and computer topics and contributes to several genealogical publications, including Ancestry Magazine and Genealogical Computing. You can e-mail him at [email protected] or visit his website at:, but he regrets that he is unable to assist with personal research.

Michael will be at the following upcoming event:

— August 18 and 19, Galesburg, Illinois. Michael is presenting a two-day workshop on using Genline for Swedish research at Carl Sandburg College. More information is at:

A printer-friendly version of this article can be found in the Library.


5 thoughts on “Not a Primary or Secondary Source, by Michael John Neill

  1. Your distinctions of primary vs. original, and secondary vs. derivative are interesting. The concept of primary and secondary sources is reinforced by genealogy software. FTM, for example, provides a space in source description where you can indicate if the source is primary or secondary. As a beginner, I used it, but stopped doing so when I realized that many sources contain both original and derivative data and therefore cannot be neatly categorized.

  2. The article was very interesting, and I now have a better appreciatipn for the distinctions between primary, secondary, original and derivative sources.

    How would you classify US census records? Or do you make that determination based on the degree of accuracy (or inaccuracy) of the data of a particular census return?

  3. I second Karen in Las Vegas’ question. In reading descriptions of how the census data was collected (sometimes from neighbors or small children if they were the only ones at home), it seems we should consider the information carefully. I’ve found family units where the birthplaces of the Head of Household’s parents change every census.

  4. There are all sorts of inaccuracies in records – many people do not remember where they were born, they remember where they lived as a child. Age too may not be accurate, even when given by the individual concerned, one of my friends was two years out when we found her birth certificate.
    Much ‘secondary’ data gives a limit – you cannot be baptised before you were born for example.
    Some inaccuracies are due to illiteracy, my illiterate gg grandmother was Arden and Harding on the same marriage certificate (if you know the accent you can see the confusion).
    Some errors are due to transcription, a lot of my ancestors who came from Antrobus were entered as coming from Autrobus in the census indexes.
    So, like evaluating most pieces of data, the researcher has a lot to think about. Because someone was there doesn’t make the information accurate – see lots of psychology experiments on observer recall. Equally because they weren’t doesn’t make it inaccurate. All researchers have to balance the information – sometimes you just have to go with the probability and too great a reliance on primary v secondary just clouds the issue.

  5. In the paragraph about the doctor being the informant for the date of death. I have a situation here that bothers me. My father died,in his home, sometime after I talked to him about 2:30 Thur. afternoon June 15 and before 5 A.M. Fri. morning. He kept a log of what time in the morning he took his pills. No times were written for Fri. The Medical Examiner(knowing the above information) put the date of death as June 17 because that is the date the body was found. Mr. Neill, do you have a comment about this kind of situation?

Leave a Reply

Your email address will not be published.