> Back to BCG Learning Center

> From OnBoard—Newsletter of BCG

Helen F. M. Leary, CG, CGL “Evidence Revisited: DNA, POE, and GPS,” OnBoard 4 (January 1998).

Science and the law are in agreement: there is only one way to prove kinships beyond reasonable doubt—DNA testing. As a genealogical standard, that is hardly practical.

For at least the past 25 years, genealogists have tried to use another standard for judging the persuasiveness of evidence: the preponderance of the evidence—popularly called the POE. In practice, the results have been troublesome.

During the summer of 1997, trustees of the Board for Certification voted to discontinue use of the term. They based that decision upon several factors discussed below.

Problem with the POE

The term preponderance of the evidence was borrowed from the legal field, where it is the standard used by judges and juries to decide the liability of a defendant in civil lawsuits. Prior to the advent of DNA testing, it was also the standard by which paternity was proved in court. One legal encyclopedia for laymen restricts it to a “case of contested facts” and defines it as “superiority in weight (determined by value and not by amount) of the evidence presented by one side over that of the other.”1

The legal standard has always been recognized as inexact when applied to genealogical compilations—a fact that has always been acknowledged by the Board, by those who lecture under its auspices, and by leading speakers and writers. It is inexact for two reasons:

  • The legal standard requires that evidence be weighty enough to only tip the scales one way or the other. In genealogy, the standard is stricter—a statement is considered “proved” when the evidence supporting it is of sufficient power to convince a reasonable, unbiased person. This does not mean “beyond the shadow of a doubt,” the strict standard used in criminal cases.2 The genealogical standard is now and always has been between the two legal extremes.
  • The genealogical standard applies when evidence is direct and virtually impregnable (e.g., an an interlocking series of birth, marriage, and death certificates, for example). It applies when circumstantial evidence is assembled—to distinguish among several persons of the same name, perhaps. And it applies when evidence is insufficient for a solid case, because records do not survive or were never created.

Although the genealogical use of POE has been beneficial in some ways—directing thought and discussion towards the nature of genealogical proof—it has often been misunderstood. Some believed it to be a method for deciding what kind of information to collect, a means of judging its probable reliability, or an arbitrary yardstick to measure “how much evidence” constitutes “proof.”

Even the leading genealogists have disagreed, over the years, as to the term’s precise meaning—although they do agree on the standard itself. Further, with the advent of televised court cases and interminable “expert” discussions of the latest Trial of the Century, legal terminology has been bandied about to the confusion of all.

Clearly, the term preponderance of the evidence has outlived its usefulness. In abandoning the term, however, the Board has not abandoned the standard—nor has the Board revised it, watered it down, or made it stricter. Perhaps it is time we called our standard by its own name: the genealogical proof standard—GPS, if we must have an acronym.

In any event, both genealogical standards and Board requirements call for an understanding of information, evidence, and proof.

Information and Evidence

DNA testing has been effectively applied in rare genealogical-historical cases—e.g., to prove that Anna Anderson a.k.a “Anastasia” was not the Russian princess of that name. DNA testing, however, is not available for general use—exhuming the bodies of all known John Smiths until an ancestral match is made is impractical, to say the least.

Because we cannot be absolutely certain of genealogical facts without the aid of DNA tests, the statements we make are assertions of what we believe to be true. If we do not safeguard the validity of our assertions, we may spend time and energy creating a hybrid—grafting somebody else’s ancestry onto our family tree. Therefore, we base every genealogical statement on evidence.

Evidence is drawn from information in written records, oral testimony, and artifacts. Information is inert and impartial, until we use it to support an assertion about someone’s identity, relationship, biographical event, or circumstance. As an example, let us say: according to the 1850 census, John M. owned real estate worth $25,900. That information is neutral—it conveys no more than exactly what it says. When we interpret the information and assert that “John M. was a wealthy landowner,” we have drawn a conclusion. That conclusion is the sum of prior deductions based on our judgment that (a) the census data is reasonably reliable; and (b) $25,900 represents realty of higher-than-normal value.3 By resting our statement on hitherto neutral information, we convert the data into evidence.

We measure the power of evidence—its convincing weight—by asking of it the following questions:

  • Are these sources generally accepted as reliable? Is the information itself coherent and free of obvious error? There are, of course, no “irrefutable” sources. Even a birth certificate may be mistaken in some of its details.
  • Was the data’s significance understood? Were its logical, physiological, psychological, social, legal, religious, ethnic, and historical contexts accurately reflected in the evidentiary conclusion?
  • Is the data being used as direct or indirect (circumstantial) evidence? This judgment depends on how it is used. “John M. was a wealthy landowner” rests on direct evidence taken from the census. The same evidence is indirect when used to support a complex statement about John’s identity—e.g., “the land value points to his being the plantation-owning John, not the Baptist minister John.”

We use these standards for negative assertions, also. “John M., who died in 1842, was not the father of Henry M., born in 1847” sounds incontrovertible but nevertheless fails if evidence about John’s death or Henry’s birth is in error.


Proof is dependent on the sum of our evidence. No single item, regardless of how reliable, well-interpreted, or direct, can stand alone in support of an assertion about a specific person.

We cannot rest a statement such as “our ancestor John M. was a wealthy landowner” on 1850 census evidence, for it does not speak to John’s identity as “our ancestor.” The sum of the evidence might include the presence in his household of Henry M., aged four — but it must also include at least two other things:

  • an evidentiary chain that links this child with the Henry M. who later fathered a child Marmaduke, who fathered a child Henry, and so on until the chain connects to us.
  • supporting evidence that the householder, John M., was indeed Henry’s father, for the enumerator in that year recorded no direct information about relationships.

Whatever the content of a genealogical statement, when we correlate and assemble a body of evidence in its support, we build a case for its validity.

Judgment about a proof’s power to convince depends on answers to another set of questions:

  • Was the search for relevant information thorough? Did it include all sources that a reasonably knowledgeable genealogist would expect of a competent search?
  • Is the evidence valid? (We return to the evaluation described above to answer this question.)
  • Was the evidence assembled correctly? Does all the underlying data concern the same person, event, or relationship? Has it been faithfully portrayed, without distortion or significant omission?
  • Does the proof satisfactorily answer questions raised by (a) data that conflicts with the genealogical statement; and/or (b) information that might have been available in records that were never created or have been subsequently destroyed.

If the answers to these questions warrant our conviction that the statement is substantially true, it is said to be proved. That does not mean that it can never be disproved by additional data or more precise interpretation. It does mean that frivolous challenges, such as “but something else could have happened” or “The History of the M. Family (undocumented) says otherwise,” should not be viewed as sufficiently powerful to destroy our case.

A lineage, pedigree, or genealogy is a series of related assertions proved by a complex web of interwoven evidence.4 In weaving that web, we cite our sources and, when necessary for clarity, explain our evidence. We may have inadvertently chosen faulty materials or made a mistaken connection, but our goal is a portrait of the family that is as close to the true picture as possible.


1. You and the Law (Pleasantville, N.Y.: Reader’s Digest, 1984), 790.

2. Technical terms for the standard used to convict in a criminal trial are clear and convincing evidence in some jurisdictions and beyond a reasonable doubt in others. Although the names themselves do not connote a restrictive standard, jury instructions normally expound on them, emphasizing the serious nature of a decision that may lead to defendant’s loss of liberty or life.

3. To make this conclusion, we would have to compare John M.’s property value with that of a reasonably broad cross-section of other landowners in the same area and time period.

4. The Board defines genealogy as a written account of the descendants of a given person or couple. Some call this kind of compilation a descendancy.


Helen F. M. Leary, CG, CGL

This article was originally published in OnBoard, BCG’s educational newsletter and is protected by copyright. Individuals may download and print copies for their personal study. Educators are granted permission to provide copies to their students as long as BCG, OnBoard, and the appropriate author are credited as the source of the material. Republication elsewhere is not permitted.