JH McConnell

What is a “non-paternity event” (NPE)? Sometimes a Y-DNA test can show unexpected results. The International Society of Genetic Genealogy (ISOGG) defines a “non-paternity event” as follows:

“Non-paternity event is a term used in genetic genealogy to describe any event which has caused a break in the link between an hereditary surname and the Y-chromosome resulting in a son using a different surname from that of his biological father. The definition excludes minor changes in the spelling of the surname, and is implicitly limited to events after the relevant branch of the surname became hereditary.”

In other words, the term “non-paternity event” can be used broadly. But in the context of this article, we’ll consider the following type of “non-paternity event”: where genetic (DNA) testing provides results that contradict a written patrilineal lineage (patrilineal meaning father to son). This implies that somewhere in the line a father was misattributed, that he was not the biological father of the next person in the lineage. Does this make sense?

Although this article will use the term “non-paternity event” for the sake of familiarity, Professor Jim Owston has proposed that the term “exogenous ancestry” would be more appropriate to the definition. If the use of this term becomes more widespread in the future, it may eliminate some of the confusion resulting from the term “non-paternity event” being used across other industries such of anthropology and sociology.

What Causes a “Non-Paternity Event”?

Two of many potential causes of a non-paternity event (in the context of genealogy) are adoptions and extra-marital pregnancy (by a man other than the husband). However, with regards to the latter, it is important to remember that each case was different. A variety of situations could lead to an extra-marital pregnancy. And not all were controlled by the mother.

DNA tests base their results on genetic similarities, a method which is usually reliable due to the complexity of genetic make-up. It is therefore highly unlikely that a close Y-DNA match will be of pure coincidence. However, as genetic genealogist Maurice Gleeson notes, we can not always rule out the possibility of a distant convergence in the results.

Why Does Any of This Matter to Genealogy?

Although the probability of finding a non-paternity event for any given ancestor is relatively low, as we will see, the percentages can add up over many generations. Think of a ten-name lineage as a chain. One missing link can break an otherwise-strong chain. Generally, the longer the lineage, the more likely a non-paternity event occurred somewhere.

Mathematical calculations by Dr. Larkin show that for the average pedigree chart (where nothing has already been DNA-confirmed), the chance of every great-great-grandfather being attributed correctly is only 73.9%. Extending four generations further, the probability is almost nil. Statistically, at least one forefather was not really a biological ancestor. However, as noted by Dr. Larkin, there is no guarantee that these averages will apply to every single family tree.

How Common Are Non-Paternity Events in Genealogy?

There are differing estimates as to historical percentages. Variants could include the geographical location, the social status of the family, and even a couple’s age at marriage, to name just three.

The International Society of Genetic Genealogy (ISOGG), quoted earlier, provides a useful list. Some estimates are specific to a geographical location. But the following two are more broad in scope, and both apply to patrilineal lineages:

• In 2009, King and Jobling accepted an overall estimate of 2% per generation for their simulation modelling, based on the probability that the actual percentages would lie between 1% and 4.5%
• Family Tree DNA (FTDNA) accept a percentage of approximately 1-2% per generation

So if we were to accept 2% as a reasonable estimate, what does this mean for a lineage (patrilineal) that spans, say, ten generations? If we were otherwise 95% certain of its complete accuracy, the 2%-based model would bring our 95% down exponentially, to around 79.2%!

Non-paternity events focus on the father, not the mother. So where a lineage spans across 15 generations but includes only 9 males (not including yourself), the calculations are similar to the example we just used (a patrilineal lineage with 10 males, or 9 not including yourself).

To calculate the above, I used this formula: P x 0.98(M). “P” is the pre-supposed probability (converted into a percentage) that the entire lineage is correct. “M” is the number of males in the lineage, not including yourself. So in the example above, the formula is: 95 x 0.98(9). You can substitute these numbers as appropriate to your equation.

As aforementioned, the true figures can differ depending on a variety of factors, including a family’s social status and culture. This article aims to give just a general gist of the overall percentages.

In summary:

• In a patrilineal lineage, the likelihood of a non-paternity event may average approximately 2% per generation, not taking into account other variables
• These calculations affect the accuracy of male lineages considerably (losing 2% per generation, exponentially)
• In an average pedigree chart extending 9 generations (including the principal and assuming nothing has already been confirmed by DNA), it is almost certain that at least one forefather has been misattributed