Understanding Genealogical Sources

Photograph by Wendy van Zyl on Pexels.com

Once I studied historical past in school I used to be launched to the idea of sources. There have been precisely three varieties:

Major Supply

“An artifact, doc, diary, manuscript, autobiography, recording, or another supply of knowledge that was created on the time underneath research. It serves as an authentic supply of details about the subject.”

Major supply, Wikipedia, https://en.wikipedia.org/wiki/Primary_source, accessed twenty third August 2020.

Secondary Supply

“A doc or recording that relates or discusses data initially introduced elsewhere.”

Secondary supply, Wikipedia, https://en.wikipedia.org/wiki/Secondary_source, accessed twenty third August 2020.

Tertiary Supply

“An index or textual consolidation of main and secondary sources.”

Tertiary supply, Wikipedia, https://en.wikipedia.org/wiki/Tertiary_source, accessed twenty third August 2020.

I neglect the precise wording utilized by my lecturers, however the variations above from Wikipedia are shut sufficient to how I keep in mind them.

However once I take into consideration sources in family tree, the definitions I discovered at school historical past classes are of restricted use. As household historians our preoccupation is with the reliability of the knowledge we discover in sources and the way it pertains to our analysis query.

Only one supply – a start, marriage or loss of life certificates, say, or a web page from a census return – can maintain a number of gadgets of details about a number of people, contributed by a number of informants, every informant in possession of various levels of data, reminiscence or written reference, reported with various levels of honesty and exactitude, maybe at an emotional second of their lives (I’ve hardly ever been as nervous and emotional as my wedding ceremony day!). The reported data was captured on paper by a clerk, registrar or census enumerator who made greatest efforts to interpret the verbal or written submissions and document the end result to the most effective of their capability, with handwriting of various legibility. The ensuing doc is accessed by us, the researcher, as an authentic of presumably compromised situation, as a picture of insufficient constancy, as a transcript of maybe unsure high quality or possibly as an index constructed upon a now difficult-to-access transcript or authentic.

These single phrase descriptions, Major, Secondary and Tertiary, appear insufficient to explain the detailed set of dependencies between the unique submission of knowledge and the attainable final result accessible to the fashionable researcher. They’ll barely assist us in any respect within the all-important dedication of reliability.

I’m not the primary genealogist to consider this downside. One among my genealogical heroes, Elizabeth Proven Mills, has devised a extra advanced mannequin for assessing genealogical sources, which she describes right here: https://www.historicpathways.com/download/templateforee.pdf

ESM’s mannequin isn’t easy, nor ought to or not it’s – no system which goals to adequately interpret and classify the complexity of the sources that we rely on can itself be trivial. Her colleagues within the Board for Certification of Genealogists within the USA have formally adopted her mannequin. The model I describe on this article is the freshest at time of writing and is described within the second version of the BCG’s Family tree Requirements.

Sorting Sources

Photograph by Ekrulila on Pexels.com

Beneath the brand new mannequin, sources are thought-about as a container for data. Our first step is to evaluate the supply. We think about the entire container, not the person data components inside it, and take into consideration what the unique creator meant. Was it:

  1. An authentic document
    Supposed by the creator as a primary recording of the knowledge, no matter whether or not this was on the time of the occasion or a while, maybe years, afterwards.
    Examples embrace:
    • Civil registrations of births, marriages and deaths captured on the authentic register workplace
    • Parish registers and banns books
    • Marriage licences and bonds
    • 1911 census family schedules
    • Wills and Administrations
    • Navy service information
    • Apprenticeship indentures
    • Passenger lists
    • Faculty registers
    • Electoral registers
    • Taxation information
    • Court docket information
    • Prison information
    • Naturalisation information
    • Land and property transaction information
    • Manorial rolls
    • Modern images and portraits
    • Gravestones
  2. A spinoff document
    A transmutation of an authentic document, whether or not that be an extract, an summary, a transcription, a translation, an index, a picture with hidden or modified data, and so on. A spinoff document should in some unspecified time in the future have been based mostly on an authentic document or a earlier spinoff of an authentic document, and we should always at all times think about what the unique document was regarded as. Word: even when the unique document underpinning a spinoff has been misplaced or broken or is tough to entry, a spinoff stays spinoff!
    Examples embrace:
    • All pictures of authentic information that lack adequate constancy or decision, or that fail to point out the entire web page, or that fail to establish the situation of the imaged web page throughout the supply, or that fail to point out the rear face of an authentic doc even when clean.
    • Civil registrations of births, marriages and deaths copied to the Common Register Workplace (transcribed from authentic registers)
    • Bishop’s transcripts of parish registers (transcribed from authentic registers)
    • All parish register indexes (together with the entire of the IGI)
    • Census enumeration books for the 1841-1901 censuses (transcribed from authentic family schedules, since destroyed)
    • Nationwide probate calendars (abstracted from authentic wills and administrations)
    • All memorial inscription transcriptions (by definition, transcribed from the unique gravestones)
    • Ballot books (derived from electoral registers)
    • Directories (commerce, avenue, postal and phone – revealed from compiled knowledge which was typically old-fashioned)
    • All translations from the unique language
    • All transcribed information on all family tree websites!
  3. An authored or narrative work
    Attracts content material mainly from authentic and spinoff information.
    Examples embrace:
    • Obituaries
    • All newspaper reviews
    • Highschool yearbooks
    • Household letters
    • Oral testimony and household tales
    • All pedigrees, household histories, memoirs, articles and analysis reviews
    • All on-line timber!

Whereas I’ve tried as an instance every supply kind with vary of examples, it may possibly generally be difficult to definitively categorise some sources. A household bible, for instance, may need the attributes of an authentic document within the fingers of an assiduous recorder; within the fingers of a much less disciplined recorder, nevertheless, it could be spinoff and even, authored.

A key level to contemplate is that authored works ought to level to the sources, whether or not authentic or spinoff, they used to attract their conclusions, and spinoff sources ought to level to the unique supply from which the info was abstracted/transcribed/listed/translated and so on. The genealogist ought to use this signposting to discover a supply as near the unique as attainable in order that they will assess it for themselves.

There’s an implicit warning right here – if the authored or spinoff supply fails to adequately establish the unique supply(s) then the researcher ought to hesitate to make use of it. The extra one makes use of authentic information, the extra dependable one’s proof shall be and the extra convincing one’s conclusions. The strongest genealogical proofs typically rely (virtually) totally upon authentic information.

Extracting and Classifying Data

Photograph by Pixabay on Pexels.com

As soon as the supply – the container – has been categorised, one can extract and classify the knowledge inside it.

Everyone knows that most of the information we rely on comprise a number of gadgets of knowledge. The step of unpacking the totally different data gadgets, reasonably than contemplating the supply as a monolithic entire, is essential to understanding the reliability of every aspect and subsequently the weights we are able to apply to them in our evaluation.

As soon as every merchandise has been extracted it’s labeled as one among three varieties:

  1. Major – first hand
  2. Secondary – second hand
  3. Undetermined

At first sight this seems easy. However how can we resolve if the knowledge supplied was given first hand or second hand? For this we’d like some data of the document kind underneath dialogue, how the knowledge was collected and underneath what circumstances.

How can we deal with main, secondary and undetermined data? A easy rule of thumb is which you could have extra confidence in main data that secondary data. Secondary data ought to at all times be corroborated earlier than it may be used with a excessive degree of confidence.

I’ll illustrate this with a few labored examples.

Instance 1: Demise Certificates of James Isherwood, 1860

This instance would, underneath the previous classification, be given the label of “Major supply.”

Deaths (CR). England. Daresbury, Cheshire. 20 Jul 1860. ISHERWOOD, James. Runcorn 8A/113.
Picture from licensed copy in possession of the creator.

We are able to extract the knowledge gadgets utilizing a easy desk:

Data extracted from loss of life certificates of James ISHERWOOD, dated 20 Jul 1860, with every merchandise categorised.

Though it’s a licensed copy from the unique register workplace, the (over) useful employees have transcribed the unique certificates reasonably than offering a picture of the unique, which might’ve been my choice. The doc is subsequently spinoff.

Assessing the person data gadgets, we are able to say that any data provided by the registrar is main (first hand). Data supplied by the informant is each main and secondary (second hand). We all know that the informant, Daniel Isherwood, was current on the loss of life, so particulars of the date and place of loss of life are main. Equally, data supplied by the informant about himself is main. Nevertheless, details about age and occupation of the deceased are almost definitely to be secondary. How do we all know whether or not Daniel knew for positive that the deceased was 62 years previous and a grasp shoemaker? The reason for loss of life, we’re advised, is licensed, in order that will need to have been supplied in writing by a health care provider and handed to the registrar, who transcribed it, so this too is main.

It’s only by consulting different sources that I could make additional evaluation of the reliability of the secondary data supplied by Daniel. From these I do know that Daniel was in reality the eldest son of James Isherwood and a shoemaker, having been apprenticed underneath his father. So, by cross-referencing I can improve the assertion of James’s occupation from secondary to main. The acknowledged age at loss of life did in reality develop into appropriate as James died lower than two weeks after his sixty-second birthday. However whether or not James ever celebrated his birthday and whether or not Daniel was ever conscious of the date of his father’s start stay unknown, so I nonetheless can’t improve that data merchandise from secondary.

Instance 2: Transcript of marriage of Daniel Isherwood and Elizabeth Allen, 1856

This instance would, underneath the previous classification, be given the label “Secondary supply.”

4 years earlier than the loss of life of James Isherwood, his son Daniel married at St Elphin’s church in Warrington, Lancashire. It’s a curse and a blessing with a city like Warrington which straddled a county boundary that some occasions of the Isherwood household occurred simply to the south of the city in Cheshire while others occurred in the principle city itself, which was in Lancashire. Whereas pictures of parish registers are broadly accessible for Cheshire occasions, these going down in Lancashire are much less effectively supported by good high quality on-line pictures. Nevertheless, there’s a thriving and essential On-line Parish Clerks venture for Lancashire, which has glorious protection for Warrington (http://www.lan-opc.org.uk/).

A fast search of the Lancashire On-line Parish Clerk web site returns a transcript of the wedding register entry for Daniel and Elizabeth:

Marriages index (PR). England. St Elphin, Warrington, Lancashire. 25 Mar 1856. ISHERWOOD, Daniel and ALLEN, Elizabeth. Web page 187, Entry 374. http://www.lan-opc.org.uk.

As earlier than, we are able to extract the knowledge utilizing a easy desk:

Data extracted from transcript of marriage of Daniel ISHERWOOD and Elizabeth ALLEN, dated 25 Mar 1856, with every merchandise categorised.

The supply is a transcript and is subsequently categorised as spinoff. The knowledge gadgets would have been contributed by no fewer than 5 totally different folks: the parish clerk, the groom, the bride and two witnesses. The clerk’s data is all main, the bride and groom’s data given about themselves is main, as are the declared names of the witnesses. However when the bride and groom provided details about their fathers, this was secondary (second hand).

By cross referencing with different sources, together with the loss of life certificates of James Isherwood above, I’m fairly positive that the knowledge supplied concerning the groom’s father is totally appropriate. Nevertheless, regardless of years of looking out, I’ve but to discover a convincing match in the precise time window for the start or baptism of an Elizabeth Allen to a father named Samuel Allen. I believe subsequently that this secondary data might in reality be technically described as a “lie”. My as but unproven speculation is that the title Samuel Allen is almost definitely an invention – Elizabeth might effectively have been illegitimate and, like so many related younger girls, merely didn’t need to admit to it in church on her wedding ceremony day.

[Aside: allowing for this being a transcription of the original, I later ordered a copy of the original marriage certificate to make sure that the details of the bride’s father had been captured correctly by Lancashire OPC. They had.]


Merely making use of a single label to a supply of “main”, “secondary” and “tertiary” is okay for a historian who’s coping with historic occasions on the macro scale. However for the household historian, coping with historical past on the micro, private degree, it isn’t sufficient.

One should individually assess the supply – the container of knowledge – after which every particular person data merchandise inside it to contemplate who contributed it, underneath what circumstances, to whom, classifying every merchandise individually as main, secondary or undetermined. Major data is of upper high quality and requires much less corroboration than secondary or undetermined data.

I hope you discovered this text helpful. Please let me know your ideas within the feedback beneath.


  1. Elizabeth Proven Mills, A Template for Evaluating Proof, Genealogical Computing 24 (Apr-June 2004), https://www.historicpathways.com/download/templateforee.pdf
  2. Family tree Requirements, Board for Certification of Genealogists (Washington DC, Second Version 2019).
  3. Elizabeth Proven Mills Ed., Skilled Family tree, Genealogical Publishing Firm (Baltimore, Maryland, 2018).