Researching the Sawtells/Dolmans of Aller, Somerset

It had been my intention for a while to try and sort out the early Sawtell family in Aller, Somerset. The period in question is the second half of the 16th century through the SCA’s grey period. I’d picked up an orphaned profile for one of them, Thomas, a while back, but hadn’t done anything. In April it had an update, which prompted me to take a closer look.

I’m not going to reproduce the derived genealogy here: you can find the current genealogy itself, with sources, on WikiTree starting with John ( WikiTree is often a collaborative effort, and it was Traci Thiessen’s initial update that brought me back to this. The timelines are mine, as are linked will transcriptions; the biographies are often a shared effort with Trudi derived from the timelines. Trudi doesn’t have access to the UK records, so was reliant initially just on Threlfall (of whom more later). It’s worth noting that, as with most genealogies of this time, it is still a work in progress.

I am going to cover some of the pitfalls and useful nuggets for SCA use.


  1. You will be cursed a thousand times by your descendants if you call your child simply John, William, Ann or Mary, or their modern equivalent. You have no excuse not to use middle names.
  2. You will be cursed a thousand times a thousand if you do it when someone else with the same surname, especially your brother, has already used the same name for his child.
  3. Do not take previously published research as automatically correct or comprehensive. Today we have far easier access to photographic reproductions of primary sources than our predecessors. Amd that applies tp new stuff dug out of the ground too. Our work won’t be the last word, because the next generation will have it even better.
  4. Be aware of context. Even apparently obvious words such as brother and sister may not mean what you initially think they do.
  5. Research around your immediate target. What might seem an obvious link and evidence, may turn out to refer to someone ro something else entirely. The same name might well actually be two distinct people. Become too focused and you’ll miss the clues.
  6. If you can’t decide between surnames, use ‘X alias Y’ and have them both.
  7. Write neatly in good ink and be praised through the centuries.

Parish Registers

It was legally mandated that parishes had to keep registers of all baptisms, marriages and burials from 1538, and that a certified copy be sent off to the Bishop, the Bishop’s Transcript, every year (at this point New Year was Lady Day, 26th March). This wasn’t always done, and when it was the quality is variable, with the best recording relationships, and very occasionally occupations, and the worst being just a dated list of names. In theory, we should have nearly half a millennium of records covering almost everyone. In practice, it is rare that any records survive all the way back to 1538, and quite common for them only to survive from around the time of the Civil War, a century later. The transcripts cover some gaps in the registers, but being a single sheet of paper sent in every year, were frequently mislaid.

With Aller, we are lucky. Not quite 1538 lucky, but the registers survive from 1560, most likely with the second register book. In addition, this register book was written neatly, in a good hand and in good ink, which hasn’t faded significantly, and with very little external damage. It is easy to read.

On a practical note, images of the Somerset registers are available via Ancestry. Unfortunately, the Ancestry indexes are often dire, especially the further back in time you go. Aller is no exception, and were clearly created by people who couldn’t read the script, or apparently have much knowledge of English names. This can be worked around, by using the indexes of the FreeREG project (but which doesn’t give access to the registers and all the details in them) and then searching for the relevant images on Ancestry. As it happened, I’d also extracted all the relevant records a couple of decades ago, back when the records were only available on microfilm, but never worked further on the early ones. Indeed, extracting the records yourself, will also give you an overall impression of the parish register, in terms of what details are recorded when, what names are locally popular, and in picking up on various family branches which might confuse identifications.

Data Bias

There’s sometimes a tendency to focus on the surviving records and forget about what got lost. This can happen whatever the subject being studied, sometimes manifesting in odd ways, such as the impression that new styles or technologies apparently sprang fully formed into existence simply because of the first surviving record or example is. The same phenomenon manifests with genealogy, often with the result of overly restricting names and families, or weird geographical jumps because of a coincident name. Hence, the Sawtells quite often get referred to as being originally from Aller. This arises out of the fact that the neighbouring parish registers did not survive as well, and so the name only tends to appear in indexes in Aller. We know the Sawtell name was more widespread locally due to other surviving records that are more inaccessible and sparser. However, whilst some members remain unaccounted for, most are, simply because geographical movement was so much less than today. Usually. We’ll come back to this.

Aliased Surnames

It’s not common, but some families used two surnames, and in the case of this family, they appear in the records interchangeably, or together. Whilst Sawtell or Sawtle is dominant, the other, Dolman, is regularly cited in combination or alone. When used together the format is always ‘X alias Y’; there are no modern forms, such as double-barreled hyphens. For the SCA, is this a valid way of having two suranmes and using them as you wish.

It wasn’t even unique to this family in Aller. One of them married a Thomas Adams (not an ancestor) alias Andrewes. The latter surname ended up in a will as Androse.

In this case, the Sawtell alias Dolman is important, as it largely only appears in Aller, and is as far as anyone can tell, certainly unique to Aller and its immediate area. This is where the data bias comes in: Aller is by far the best recorded parish. So whilst what comes matches up with Aller records, it cannot be ruled out that we’re all being fooled and an entirely different branch of the family from across the river, at Huish Episcopi, for example, is actually involved. Why? Because whilst Sawtell alias Dolman is localised, given names such as John, William, Thomas, Agnes and Joan are not, and even within Aller we have two brothers, John and William, using all those names for their children, baptised during the same couple of decades. So even when the surname is localised, how do you tell one John from another? And how do you know if the record of a third has not survived at all?

There are clues, some obvious, such as ‘wife of’, or ‘son of’, and others which are a balance of probability such as it being more likely that two siblings will be buried within a day of each other, rather than cousins. But still only a probability, not a definite.

The Perils of Emigration

This is where the aforesaid word ‘usually’ comes in. It turns out that (with due regard to the data bias caveat) that three of this family, of whichever line, emigrated to New England and New York, probably in the 1610s or 1620s. This, of course, was a perilous journey into a dangerous land (often because the existing inhabitants didn’t regard it as empty), but all three seem to have ended up doing pretty well. But this isn’t the peril I refer to.

The peril is that of the gateway ancestor, to wherever or whenever that may be. In this case, early immigrants to the New World are desirable members of the family, and tend to excite a lot of interest. Now we have three of them, most likely from Aller, and hence siblings or first cousins. This leads us to

The Peril of Printed Genealogies

This is a problem that has existed since the SCA period: we only have to look at genealogies that end in various divine pantheons or refugees from Troy. The problem extends forwards, not just with explicitly genealogical publications, such as visitations, but references such as Burke’s Peerage or lists of students at universities. The problem tends to be that genealogies, in general, do not elicit a great deal of academic or other research (with exceptions, usually centred on powerful families), and this often means that whoever publishes first often publishes alone, and is forever copied. With older references, of which the Victorians were quite fond, this often arose either from now obvious research errors or sanitisation of family history by the (usually) gentry concerned. “No-one will notice if we say we got married a year earlier than we did, and just a month before our eldest son was baptised, will they?”

The problem isn’t usually one of outright deceit though, nor lack of skill. Sometimes there can be an obvious lack of care or time. But all of this comes down in the end to the logistics of conducting the research. Parish registers were largely kept by the individual parish churches until 1978, which meant visitmg each individually to see the original. One could rely on transcripts, such as those by the Society of Genealogists, but they were a limited subset. After 1978 the registers had to be deposited at the county archives or record offices. That at least meant you could see all the registers for a given county in a single location.

Then the Mormons, or the CHuurch of Latter Day Saints came along, driven by their particular theology, starting the International Genealogical Index (IGI), which is now, after some persuasion by a few of us in the College of Heralds back in the early 2010s, widely used in name registrations within the SCA. It’s not pergfect, but far better than the previous limitation of a small number of older books. These days, we can bypass the IGI and go straight to the registers themselves, but for many that is overkill. The LDS also started a widespread operation of microfilming the registers, which could also be purchased from the archives at a fair cost, but generally less than that required to visit a far flung archive. And for the first time there was a central location for all these filmed records; unfortunately for us in England, it was in Salt Lake City. I still have many parish registers on microfilm or fiche, and did once take a couple of days in Salt Lake City, before I joined the SCA, to visit the LDS archive. But this was still work in progress in the early 21st century.

With this brief background, it is possible to see the effort required for a past genealogist to see just the parish registers (and we haven’t touched on items such as wills). Now, I can browse through almost all counties on any device for a subscription, whenever and wherever I have time. In passing it is worth noting that the same limitations apply to a lot of the older printed dictionaries of personal and place names used for SCA registrations. Whilst these sources are accurate in their citations, they are nowhere near comprehensive and lack of evidence within them should not be taken as evidence of a name not being used. You may just need to look elsewhere.

The Perils Combined

Pulling the focus back to the Sawtell alias Dolmans, the published book in question is ‘Fifty Great Migration Colonists to New England & Their Origins’ by John Brookes Threlfall, published in 1990.

For the Sawtells, Threlfall starts with Richard, one of the immigrants, baptised in Aller in 1611. One of the complications with the Aller family is that at this point there were two brothers, John and William, having children with a hefty overlap in time and names, mostly with very little imagination. Unfortunately, Threlfall largely ignored William, which left him blind to certain options. However, only John named a son Richard, so we can at least be confident about his parentage. Of his life in New England I can’t comment: I’m not familiar with the (nw) U.S. records and his emigration removes him from consideration in the English lines.

Threlfall then moves on to his father, John, and we start running into problems. The first, from an SCA registration point of view, is that Threlfall has normalised names, and starts by swapping John’s wife, Agnes’s, maiden name to Pittard from Pyttard. The name only comes from one source, so in this case particularly, it seems an unnecessary change.

The next problem is that of the deaths or, more precisely, the burials of John and Agnes. Threlfall came across the burial of John and Agnes a day apart in 1622, and ran straight into a common pitfall with genealogy, namely coincidence. I can only conclude that he saw this pair of burials, assumed they must relate to the couple and then stopped. However, if you carry on, in January 1635/6, you come across the burial, again a day apart, of Agnes and John. This time they are identified beyond just names, with Agnes being explicitly described as the wife of John, and John described as ‘thelder’. Threlfall missed this pair, and in doing so didn’t consider who else could have been buried. As was common until recently, John and Agnes had named two of their children after themselves, and it is these that died in 1622. This in turn allows us to narrow down who was still available to be recorded in other later records, such as marriages and wills. This is especially helpful in a village that would have otherwise had two Johns of very similar age (just one of their pairs).

Threlfall ends his biography of John with a list of his alleged nine children, the first baptised in 1600. Unfortunately for Threlfall, there are actually eleven, and one of the ones he does have is completely spurious. This is where past errors, combined with lack of context, start to magnify.

The lack of context is that Threlfall has failed to do much in the way of finding out what cousins the children of John had. This has left him blind to alternative cousins of about the same age that could also be candidates for certain events.

For example, Threlfall is confused about the Henry who ended up in New York, as he is unaware of any living. He has one as the son of John, who died as an infant, but has missed John’s second Henry, and also his first cousin Henry, son of William, baptised in 1609 amd 1610. So now we have gone from none to two, and it is still unclear which is the likely candidate, but we can no longer proceed on the assumption that it was just siblings that emigrated to the New World.

A second error is that Threlfall has John’s eldest sonm also John, marring a Katherine (still of unknown origin). However, (this is where keeping track distinct individuals gets hard) we now know John son of John ‘thelder’ died in 1622 before this marriage. What Threlfall apparently didn’t know is that John ‘thelder’’s brother, Willliam, also had a son John, who we know was still alive in 1629, thanks to a will. This means that Threlfall has assigned the marriage to the wrong, dead cousin.

The most interesting one though, is the spurious child, Ann. There is no baptism for Ann, and it might be thought that she is actually Agnes; I have seen the names used interchangeably a couple of times. However, this isn’t the case here. So where did Threlfall conjure her from? The source is Richard’s probable brother Thomas’s will in New England, in which he references his brother Richard Sawtell and sister Kenricke of Muddy River. It turns out that this was a married lady called Ann Kenricke, assumed by Threlfall to be originally a Sawtell. In usual English wills this formulation would most likely mean that Ann was indeed his biological sister.

However, this ignores the context of the will and those in it. Further research, not mine, shows that Ann was actually an Ann Smith, not Sawtell. So what is going on? The will is not in actual fact a written will by Thomas, but an attestation of Thomas’s wishes testified by his Presbyterian minister. The whole community of immigrants was staunchly protestant, and had emigrated in order to be able to gain their religious freedom. It was often their practice to refer to anyone else in their community as brother or sister. Thus when the minister spoke of sister Kenricke, he was not referring to a biological relationship. This also reopens the question of who exactly Thomas was with respect to Richard. Richard did have a biological brother Thomas, but he also had a first cousin Thomas as well, both of whom seem to vanish from the English records after their baptisms.

Going back further in time is hard. We know John ‘thelder’ had at least three siblings, one being the aforementioned William, thanks to their father’s, another John, will. Unfortunately, their baptisms push us beyond the start of Aller’s registers. Threlfall mentions, completely unsourced, another will from 1549, but so far I have found no other reference to it that does not lead back tp Threlfall, which is slightly worrying. I don’t believe Threlfall made it up, but he could well have picked it up from a source that did. I suspect even Threlfall found it suspect given that he uncharacteristically doesn’t give any citation.

A Final Nugget

One of William’s children, baptised in February 1605/6 was named Marmaduke. This was an early use of a given name that would become extremely popular, alongside the usual suspects, in Aller and the immediate area. This was across all classes from the high born, through the yeomanry, right down to the labourers. There is only one registered Marmaduke in the OSCAR era of the SCA, whose documentation shows the name to geographically quite widespread. But in this corner of Somerset may well be where iy became most popular.