Determining reliability is not a simple matter. The terms “primary source” and “secondary source” theoretically distinguish reliable sources from potentially unreliable ones. As a point of fact, however, in various fields the terms are used ambiguously by researchers in a number of contradictory ways. Attempting to make a simple either-or choice does not enable a researcher to evaluate historical evidence reliably.
Genealogical practice appraises reliability in three ways. We appraise the source (its physical form), the quality of the information within that source, and the type of evidence we can draw from that information. Each of those aspects has three basic qualities. The following provides a brief tutorial.
Sources can be people, artifacts, documents, or publications (printed or digital). They are either
• original records, that is, those not based on a prior record;
• derivative records, that is, records created from prior records by manipulating their content-as with abstracts, compilations, databases, extracts, transcripts, and translations; or
• authored works, that is, written products that synthesize information from many prior sources and present the writer’s own conclusions, interpretations, and thoughts.
In using a source, we evaluate separately each information statement, to determine whether it offers
• primary information, that is, details provided by someone with firsthand knowledge of the “fact” reported;
• secondary information, that is, details provided by someone with secondhand or more-distant knowledge (aka, hearsay); or
• undetermined information, that is, details provided by someone whose identity is not known.
Information that is relevant to the problem is considered evidence. It may be one of three basic types:
• direct evidence, that is, relevant information that seems to answer the research question all by itself;
• indirect evidence, that is, relevant information that cannot, alone, answer the research question; rather, it must be combined with other information to arrive at an answer; or
• negative evidence, that is, evidence arising from an absence of a situation or information in extant records where that information might be expected.
A map illustrating this evaluation process appears here.