Anders Family
David Asbel Anders with
Jim and Tina Smith Anders

US GenWeb Project

Rufus and Florence Hall Owen Family in 1947
Rufus and Florence Hall Owen and children

US GenWeb Archives Project


Transylvania County, NC 

GenWeb Project

NC GenWeb


Primary vs. Secondary Sources

Print This Page

Introducing our Contestants

You will frequently see these terms to distinguish what kind of source was used for a particular piece of information or to identify a source itself. Essentially, a Primary Source was created by someone who was there. These include diaries, letters, court records, business records, census records and countless other documents. Some occur as the event actually happens, such as a purchase receipt. Some are actually recorded after the event. For example, a doctor shouldn’t be filling out the birth certificate as the child is born. He needs to finish he job, clean up a bit, and then pull out the pen. Length of time between the event and the record varies. But in all cases the person behind the pen was there. Secondary Sources are everything else. These range from classic history texts to your personal genealogy database. Here, the person behind the pen wasn’t actually there. Instead, the individual has compiled information from various sources to complete the information.

Determining if a source is primary or secondary can be pretty tricky. For example, newspapers can be both primary and secondary sources. During the American Bicentennial, the Transylvania Times reported many local events to celebrate the occasion. The reporter was actually there, becoming the “pen” for participants to record the event. These articles are primary sources. Likewise, throughout the year, the newspaper published several articles bout the history of Transylvania County. Many of these dealt with topics for which all participants are deceased. As a result, these articles are secondary sources. A Federal Census record is considered a primary source. A transcription of that census is considered a secondary source since the transcriber has an opportunity to add additional research – hopefully clarifying which is from the document and which is from elsewhere – as well as add additional errors. Interestingly, even though the actual Federal Census microfilm is considered a primary source, these records are actually transcriptions of the state copy. This detail is generally allowed to slide since there weren’t photocopiers available at the time. Researchers should simply keep in mind that yet another opportunity for error was involved. To further complicate matters, some academic editions of transcribed manuscripts are of such good quality and contain so much additional editorial information they are "promoted" to a "Published Primary Source" in bibliographies used in academic research.

Round One - Primary Sources

Primary Sources are generally considered to be more reliable than secondary sources. But this is not always the case. People make mistakes, or their memory gets a little soft, or, on occasion, they just point blank lie. One primary source alone should never be considered as absolute proof of anything. Users should carefully consider the context of the document and the author. A birth certificate was completed by the doctor immediately following the birth. Short of deliberate fraud, we can be certain the identity of the mother and child is correct. But what about the father? The doctor wasn’t there at conception. Likewise personal memoirs of events can add new personal interpretations of events that would be far different than say a diary or letter written as it happened. Census records are notorious for errors for a multitude of reasons. Any set of application involving personal gain, such a revolutionary war grant and Cherokee enrollments will generally contain some degree of fraud. And we have all experienced getting the details about something a bit mixed up at points. Another significant factor, particularly in older newspaper accounts, is the bias of the author. In Western North Carolina newspapers at the beginning of the 20th century pretty much carried a strong Democratic Party bias. Reports of various party politics reflect this. When reading today’s news readers hopefully take these editorial biases into account. We should do the same when reading information from older sources.

Round Two - Secondary Sources

When done well, secondary sources can help evaluate primary sources by providing historic context and alternative views of events. When done poorly, secondary sources can send researchers in the wrong direction or in the worse case scenario result in poor public policy decisions concerning historic resources. Again, readers should carefully consider the source before evaluating its value to current research. Secondary Sources should include the sources for the research. Unfortunately, many beginning researchers are not aware of the importance of sources – though they tend to regret not having taken better source notes later on. Also, many publishers try to cut costs by not including source information. If enough disappointed readers write the publishers about this omission, this trend will change. Until then, this practice greatly hinders future research. The background of the author is important as well. Professional historians will have a much deeper understanding of the historic context and important issues that may be involved. This is important when dealing with larger topics. A good family historian will have much better information on a particular family and the events that shaped their lives. The source information and background of the researcher help readers evaluate the usefulness of the work. 

And the Winner is?

So which kind of source is better? Neither. Good research will always involve a combination of sources. Book reviews and annotated bibliographies contain information about the strengths and weaknesses of a work as part of this evaluation and are meant only to help place the work within the whole body of available research, not praise or condemn the author. Edited versions of primary documents add crucial information in evaluating original source material. What is crucial is to allow the often conflicting information to tell its own story. One of the most common mistakes in genealogy is to simplify the discrepancies by dismissing information as erroneous in order to support a predetermined conclusion. In Western North Carolina research as frequent example of this is the conclusion that an individual settled in one of the far western counties earlier than he actually did – assuming quirks in the early documents.   More often the ancestor actually migrated west in steps. It is easy to make such assumptions. And erroneous assumptions can lead to years of misplaced research following the wrong interpretation of events. 

We live in a complicated world filled with complicated people. Events were no less complicated in the past. Ultimately, each researcher should constantly seek all information available on a topic and then add their own research for the next researcher who joins in the never ending process of understanding our past. 

 - Linda Hoxit Raxter, originally posted January 4, 2003

^ Back to Top