Sources, documents, records, data, facts, information, evidence. They’re similar terms and we family historians sometimes use them more loosely than we should. Of all these terms, the most vital, and perhaps the least well understood, is evidence. We all have implicit notions of what constitutes evidence, but it has a precise meaning in genealogy which is not well understood. Knowing how to clearly define, classify, group and use genealogical evidence is an essential skill for all family historians.
I searched for definitions of “evidence” according to various genealogical bodies around the world. To my surprise I found only one organisation, the US-based Board of Certification for Genealogists, have a published definition:
A research question’s tentative answer, which can be right or wrong, complete or incomplete, or vague or specific; can be direct, indirect, or negative.Board for Certification of Genealogists, Genealogy Standards, Second Edition, 2019, p.77. (“BCG” hereafter.)
Hmm. While the large number of clauses in that sentence show that the authors have striven for a complete definition, it isn’t the clearest or most concise definition I’ve ever read.
The reason the BCG definition is so impenetrable is that it’s trying to summarise a sophisticated model for the derivation of evidence (which can be direct, indirect, or negative) from information (which can be primary, secondary, or undetermined), itself derived from sources (which can be original, derivative or authored), all in one sentence.
The BCG definition of evidence is built on no fewer than eleven other terms that each have their own definitions. This is not a user-friendly approach.
In this article I will focus in on the core elements of genealogical evidence and try to make them as clear and easy to use as I can.
Although poorly presented, the BCG model for genealogical evidence has a great strength at its heart: all genealogy research is based on sources, those sources contain information, and evidence is derived from that information.
It sounds self-evident, but there is some real power in stating it this way, and some hidden depths:
- Information on its own is NOT evidence. The researcher must draw the evidence from the stated or inferred information.
- A source on its own is NOT information. It is the container from which information can be extracted. As family historians we must use our knowledge of the source to correctly identify and interpret the information held within it.
Types of Evidence
There are three types of evidence in the BCG model: direct, indirect, and negative.
Let’s look at them one at a time:
- Direct Evidence
Will prove a point of fact on its own and without interpretation of circumstances.
I’m looking for the marriage of my Great-great-grandfather John Bowers. I find a marriage certificate showing he married one Mary Jane Hyland in Warrington, Lancashire in 1894. This is direct evidence of the marriage; I need no other evidence to establish this as a point of fact.
- Indirect Evidence
Will prove a point of fact only in combination with other elements of indirect evidence, i.e. it requires some interpretation of the circumstances. In legal parlance, this would be circumstantial evidence.
I’m researching the Hyland family of Liverpool, Lancashire. I suspect they may have been Irish immigrants so their religious denomination would most likely have been Catholic rather than Church of England. I find a batch baptism of the three Hyland children into the Church of England in 1869 when the children are aged nine, six and four. Batch baptisms of children of this age, while not unknown, were not standard practice. One interpretation of this information is that they may have been baptised into the Catholic church as infants and converted at a later time. Further searches find an earlier Catholic baptism for the middle child in 1863, though I’m currently unable to find Catholic baptisms of the other two children in the extant online sources. Taken together, these two pieces of information support my theory that the family were initially Catholic and later converted to the CofE. On their own, neither piece of indirect evidence shows a change of religion, it is only when combined that we can see that the change took place.
- Negative Evidence
Indirect evidence arising from the absence of information where it might otherwise be expected.
Negative evidence must be handled with care however – be mindful of the old adage that “absence of evidence does not mean evidence of absence“. It can only ever be interpreted as indirect evidence, requiring additional evidence before it can constitute proof.
I’m looking for the birth parents of Ernest Albert Stringer, born in Cheshire in 1872. I find a birth certificate for Ernest which names his mother, Anne Stringer, but no father. This is strong negative evidence that Ernest was born illegitimately. Further searches show that Anne was unmarried and aged just 19 at the time of Ernest’s birth. She subsequently remarried and left Ernest to be brought up by her parents. Combining the negative evidence of the birth certificate with the other indirect evidence, we can make a very strong case that Ernest was illegitimate.
I’m researching the Hyland family of Liverpool, Lancashire (again). I believe that William Hyland, the father of the family was dead by the time his daughter Mary Jane married in Warrington in 1894, as she stated that he was deceased on her marriage certificate. I find the family in both 1881 and 1871 censuses, but on both occasions William is absent. This is negative evidence suggesting that William probably died before 1871. Further searches reveal that William died in December 1865. In this case, the negative evidence of the marriage certificate and the censuses helped me to narrow down the search window for William’s death, leading to the direct evidence of his death.
Be completely honest with yourself, do you structure and classify the evidence you collect in this manner? I think this is a useful tool, and one that we should all be using and discussing more widely.
The Need for Source Citations
We are often told that it is critical that we cite all our sources, but there is seldom a discussion of why? The point of the source citation is to enable another researcher (or ourselves at a future time, when we’ve long forgotten the details) to re-create the Source –> Information –> Evidence path of reasoning. By shining a light on what we rely upon to derive our evidence, we enable open review, discussion and, if necessary, revision of the chain of reasoning underpinning the evidence we present towards our overall conclusions.
Having a transparent chain of reasoning is a fundamental principal in all investigative disciplines. In science, for example, researchers publish articles in peer-reviewed journals, fully describing the process they followed to achieve their results. This enables other scientists to reproduce the experiment and confirm the result. Without reproducibility, a scientific result is considered null and void, causing it to be formally withdrawn. (This can and does happen – c.f. the furore over Cold Fusion in 1989 – see https://en.wikipedia.org/wiki/Cold_fusion)
In genealogy, we must aim for similar levels of transparency in our research. So, our rule of thumb here should be: evidence is only evidence when supported by source citations.
If only all those online trees held enough information to uphold their assertions…
While the BCG model is a little daunting, it is the best model we have for genealogical evidence. It enables us to classify and collate evidence in a structured way, leading to valid conclusions that are transparent to others and will stand the test of time.
Next Time: I will look at the challenge of linking two or more pieces of evidence together. When we find two records in disparate times and places, how can we satisfy ourselves that they relate to the same individual?