Document Analysis Exercise: Parish Register



Yesterday, to rouse EE users out of the pre-holiday Friday doldrums, we posted this challenge at our Facebook page:

Document Analysis Exercise
Let’s say you have just received the attached image from someone,1 with no information except it’s a baptism of David Murison, 1850-51, in the church records of Alyth Parish (Perthshire), Scotland, held by “OPR” and copied from “FHL microfilm 993,514.” How would you handle it?

The Response
Interestingly, the first responses focused on finding the source to create a better citation. For certain, that’s important. Absolutely important. But .... what if you don’t have a costly subscription to the one online database that offers better images?2 What if you do get the original and it’s no more legible?  More to the point, what’s the difference between analysis of a record and source identification?3

Hazel and Dana were curious about the statement that David Murison was a farmer from “Cult” or “Cults”4 and suggested next steps to pursue research there. Kate suggested that we should use the record as a springboard to find the couple on the 1851 census of Scotland.  Aine’s “first instinct,” after finding that the FHL microfilm was not digitized online, was to capture the “transcription” [more precisely, extraction]5 that is offered online at FamilySearch—and posted it along with FS’s suggested citation.

All good. But, of course, next steps and finding extractions that are more readable aren’t the same as analysis.  We must analyze the content before we can decide on the appropriate next steps we can use to build upon this data. That analysis should include

  • all content observable on the page—not just the details of the one baptismal act in which we are interested.
  • correlation of details from other entries against the details of our one item. Therein lie clues that help us better understand the situation, the source we are using, and the reliability of the data.

Dana, after noting that the father was from a different village, spotlighted a key analytical issue: The scribe “did not record the baptisms in exact date order. Why?” Kate offered a suggestion: “Seems as though they were not written … at the time of occurrence but at a later date[,] maybe as notes were found.”

This kind of analysis can trigger a whole host of questions and debates:

  1. What was the year of the child’s birth—given that the page header is “1850-1” but the entries are not in chronological sequence? If we refer to the FamilySearch database entry of extracted details that Aine provided, we see that FamilySearch assigns the year 1851. On what basis?
  2. The scribe habitually cited a place of residence for each father. Some were residents of the town in which the church was situated. Some resided elsewhere. Did the parents from surrounding villages come into the parish church or did the minister go out into the hinterlands of his parish to baptize those children? Comparing all dates on the full page against a perpetual calendar is often a revealing exercise.
  3. Why are most entries in chronological sequence by date of baptism—but random exceptions appear? Does this signify that part or all the register was compiled at some later point from loose notes, a process that might result in copying errors? Logically, if the minister baptized infants at their homes in outlying villages, he would not take his register with him. However, an analysis of all the entries on that page show some out-of-sequence baptisms for Alyth itself, the site of the church. Why, in those cases, would he not record the baptisms at the time they occurred?
  4. When the entry says that the child was born "7 Feby.” and baptized “24 Ditto,” in what month did that baptism occur?  If Feby. was intended for the baptism, why did the scribe not simply repeat that abbreviation, Feby? Space is tight at the end of the line of script, but the scribe’s “Feby.” took no more space to write than did his “Ditto.”
  5. Are there other usages of “Ditto” on that page that might shed more light on the scribe’s intent? If not, should we not obtain images of surrounding pages so we have a larger sampling to study and, thereby, draw a more reliable conclusion about the baptismal date of this child?
  6. In common usage, “Ditto” might mean “as above” or “the same.” Commonly when historical researchers see “Ditto” in documents, that usage appears on tables or lists where “Ditto” means “as above in this same spot [or field or context].” If we look at the entry above David, we find that the scribe wrote there: “born 30th January and baptized.” End of line. No baptismal date given. Either the scribe was absent-minded or else he did not have a date of baptism to enter there. Did the scribe use “ditto” for David’s entry to signify “the same situation as above”—i.e., that he could not now say what month the baptism occurred?

Aine also took her search for “a better image” one step further and provided us with a new image—a two-page spread showing all entries on the preceding page.

That has answered some questions but raised others:

  1. The heading on the prior page is “1850.” The heading on our page of interest is “1850-1.” The pattern here suggests that David’s birth in February, half-way down the 1850-1 page, should have occurred in 1851.
  2. The prior page also sheds more light on the question “Were most register entries recorded at the time the baptism occurred or was this register assembled at a later date from loose records?”  The third entry on that page, for Margaret Forbes, tells us that she was “born 17th April and baptized 2nd June, though here inserted.”  Margaret’s entry appears in this sequence of baptismal dates

⦁    baptized 1st September
⦁    baptized 4th September
⦁    Margaret, baptized 2nd June
⦁    baptized 8th August
⦁    baptized 18th October
⦁    baptized 20th October
⦁    baptized 22nd October
⦁    baptized 27th October
⦁    baptized __ November
⦁    baptized 29 December
⦁    baptized 29 December
⦁    baptized 10 November 1851 [sic]
⦁    baptized 5 January

Three Last questions:
Anita also notes that “Most babies were baptized 1-3 months after their birth. Why would that matter to you in your research?

Amid your own analysis of these two pages, what other patterns, clues, or questions can you draw that haven’t yet been discussed?  We'd love to hear your other ideas in the Comment screen below.

And the most critical question of all:
We all know how often records disagree with each other. If you found another record that gave a different date of birth or baptism for David Murison, what would be your reaction? Would you say: But I have his baptismal record which gives the date as xyz. That baptismal record is indisputably reliable.

1. The Image and problem are compliments of Gary Gauthier. The photocopy that Gary holds, provided by the duplication service of the Family History Library (FHL) in 2003, is a better quality than the posted image. For his own files, he digitized the image at a high resolution (2400 dpi), creating a digital file too large to upload. To effect the upload, he downsized the pixels, which also lessened the quality slightly. For the full thread of our discussion, which included a variety of citation issues, see

Those technical details are beside the point, however. Historical researchers regularly work with records offering various stages of readability. The content of what we have at hand still has to be analyzed until we succeed in tracking down a higher quality image—if our analysis convinces us that a better (and perhaps costly) image is essential.

2.  National Records of Scotland, ScotlandsPeople (

3.  This, of course, gets to the reason why the exercise was labeled “Document Analysis Exercise” rather than “Citation Exercise.”

4. The place name is actually written as “Cull.” The Google search by our respondents turned up the town of “Cults” but not Cull. However, Cults lies in Aberdeenshire. The Parish of Alyth lay in Perthshire, according to cataloging at FHL and ScotlandsPeople. Alyth in Perthshire and Cults in Aberdeenshire are currently 67 miles apart. Should you accept the Google or Google maps search and conclude that David’s parents were from Aberdeenshire and refocus your search there? Or would you seek out histories of Perthshire or Alyth in search of a similarly named village there?

In response to the issue raised in the Facebook thread (and, coincidentally, while this blog post was being drafted), Gary answered that question for us, in the Citation Issues Forum thread he had begun, by providing a link to a Murison family page that identifies Cult as a farm outside of Alyth. As a farm, the odds of it appearing on Google maps would be slim.

5. To be precise (which researchers should, of course) the set of details that FamilySearch offers for this record is not a transcription. A transcription is an exact copy of the content of the record—one that maintains all words in their original sequence and spelling, omits no words, and adds no details or punctuation unless the addition is placed in square editorial brackets [ ].

       What FamilySearch offers is a database entry, which we might call an “extraction.” Engineers first created a database with standardized fields that would cover basic details from a wide range of sources. Data was then extracted from each record book into those preset fields. When this-or-that register offered details for which the database had no field, that data was omitted. In this case, the omission included both the occupation and place of residence of the child’s father.


Image 1:
Church of Scotland, Alyth Parish (Perthshire), Baptisms, marriages and burials, v. 6, 1820-1854, p. 68 headed “1850-1," baptism entry for David, son of David Murison and Mary Duncan, born 7 February (no year specified but apparently 1851); baptized "24 ditto" (i.e., 24th of ___, month not specified); FHL microfilm 993,514, for which the catalog description at FamilySearch ( : accessed 13 May 2019) cites "Microfilm of O.P.R. [Old Parish Registers] ms. no. 328 in the New Register House, Edinburgh"; image printout supplied 2003 by FHL’s document copy service.

Image 2:
Image provided by Aine Ni Donnghaile to Elizabeth Shown Mills by Facebook Messenger, 24 May 2019.


“Document Analysis Exercise: Parish Register,” blog post, QuickTips: The Blog @ Evidence Explained ( : posted 25 May 2019).

Submitted byHistory-Hunteron Sat, 05/25/2019 - 18:38

Dear Editor;

I noted with some interest that there was an assumption made by a reader that; “Most babies were baptized 1-3 months after their birth. Why would that matter to you in your research?"

In my research of my grandmothers' family, I just found the parish registers which show that she and three of her siblings were baptized long after the noted time period had elapsed. In fact; she was 8 years old and the family had them baptized all at the same time. Of course; this begs the question as to why. It seems that one of the boys must have been ill, as he passed away shortly after. This type of situation teaches us that assumptions are "dangerous". People do things for a reason, but not necessarily one known to us. We need to be so careful about our assertions, when it comes to genealogy.

Submitted byEEon Sat, 05/25/2019 - 20:39

H-H, when I read that statement in the Facebook observation, I took it differently. My take was that the respondent was paying careful attention to the pattern in that particular parish in that particular time frame. Definitely, baptismal patterns vary across time and place. For example, on the colonial Louisiana-Texas frontier, where I've cataloged all sorts of societal patterns, pre-1763 baptisms were almost always within 1-2 days of birth, for both the free population and the enslaved. After 1763, however, we begin to see great delays, from several months to 2-3 years. That, of course, raises a question of its own: What was the catalyst for the change?  (And yes, there was one.)

I'd say that the danger lay in applying to one locale or one time-frame a pattern one has observed elsewhere, without making the effort to define the pattern for the specific place and time.

Submitted byRobynRon Sat, 06/01/2019 - 07:03

Mary Thomson baptism 1827 Carnbee, FifeAs I research in Scotland fairly full on, but do so in spurts (my mother's side of our family), I was very interested in reading this "Document Analysis Exercise: Parish Register". I am tonight actually attempting to write a proof argument of why many researchers, are incorrectly showing the husband of so of so (Scotland). That's another story for later.

But in regards to this thread, I have another one (question re interpretation).

In 1827, a child of Hugh Thomson and Mary Herd was born on the 2d of March and baptised on the 3d June.

On "every other entry" on the same page, the parents are recorded as so & so and "his wife" - except the one of interest noted and the last entry that clearly states "both unmarried".

I believe that Hugh Thomson and Mary Herd married 31 March 1827 in Carnbee.

Is this a reason why Mary Herd was not recorded as wife when daughter Mary was baptised?