by Peter H Saul
This article was originally published in the August 2006 edition of Soul Search, the journal of The Sole Society
In addition to straightforward genealogy, I have long had an interest in the origins of the surname Saul.
An excellent introduction to the topic was given by Don Steel in Sole Search of November 1993, with a preceding article in June 1992. There are a number of theories as to the origin of the name, and in all probability, most of them are correct, i.e. the name has several origins. Those most recognised are :-
1) The Lake District Sauls, probably originating from the Norse personal name SJOLFR, a contraction of SAE-ULFR, the Sea Wolf. These Sauls are also the origin of the Sauls in Lancashire, notably around Preston, Garstang, Winmarleigh and Liverpool.
2) The Gloucestershire Sauls, named after the village of Saul. These are the origin of the Banbury Sauls too. Research here suggests that Sauls in Gloucestershire also have more than one root. This village was not noted in Domesday book, which possibly argues against Viking origin.
3) The Norfolk Sauls, named after the village of Salle (pronounced Saul). This village, as Sall, predated the Norman Conquest, and had 4 entries in the Domesday book.
4) The Norman Knight De La Salle (or Sale, or Sall or Saule, I have seen all of these variants). He was granted lands in Ireland, and may be one of the Irish sources of the Saul surname.
5) The Co. Down town of Saul is associated with St. Patrick, and is derived from the Old Irish Sabul, meaning barn. St. Patrick may be buried at Saul; or a number of other places. Saints Bridgit and Columba may also be buried there (or not; a challenging piece of history).
6) Les Saul made a case in November 1998 for pre-Norman Sauls in Hampreston, Dorset, as reported in the Domesday book. These could well be an offshoot of the Norse group above.
7) Saul was of course a Biblical name, although the pronunciation would be shawl. It remains a known given name in both Jewish and non-Jewish traditions. My own research on visits to Israel indicated that it not currently used as a family name. As this article was being completed, I did see a newspaper reference to a Yehudah Shaul in Israel, so it remains a possibility.
8) Least likely in the U.K., but a possibility in Germany, is Bad Saulgau, in Baden-Wurttemburg. The first part of this would be pronounced Sowl, and I have been asked if that is correct for my surname when in Germany.
Thus, there are possibly 8 origins of the name. Of the list above, speculatively, 6 may be an offshoot of 1, and 4 and 5 are probably not separable. Three may be geographically identified in England and one in Ireland while two (7 and 8 above) are considered relatively unlikely.
The primary Norse origin of Saul is “Sjolfr”. The origin of the name was in Old West Norse (OW.Norse in references) the language used in Norway, Iceland, and other parts of western Scandinavia. Variants found in Runic inscriptions are SæulfR, Siólfr,Sæúlfr, and Sjólfr. The first element Sæ- or Søy- is from OW.Norse sjár, sjór, sær (from proto-Scandinavian saiwaR “sea, ocean”). The second element -úlfr and the side form -ólfr are from wulfaR, “wolf”. When this second element appears in the latter part of masculine names, it is always pronounced as -ólfr, and quite often spelled that way as well. Sjólfr was a name used by some berserkers, hence the wolf connotation; some berserkers wore wolf skins instead of armour.
Some historians distinguish between the areas settled by the Norwegians and those occupied by the Danes. (Cameron (1996)). The Norwegians, both in direct settlement and by migration from Ireland especially after their expulsion in 902 were mainly established in the north-west, around Cumberland, Westmorland and Lancashire, i.e. both the part of Lancashire now in Cumbria and in the main part of Lancashire. The map illustrates the approximate settlement patterns. Loyn (1962) pointed out that there was some movement across the Pennines into the West and North Ridings of Yorkshire. Indeed, it is known that Norwegians from Dublin established a kingdom in York. The many Norwegian place-names in the area suggest major settlement there. There is also evidence of Danish settlement in parts of Westmorland and in Manchester. Nevertheless, the highest concentration of Danes occurred in the areas occupied by present-day Yorkshire, Norfolk, Leicestershire, Lincolnshire, Nottinghamshire and Derbyshire.
Since the Vikings moved at least in part from Ireland to NW England, it may never be possible to separate the Irish sources of the name; but it is probable that the Viking connection was predominant.
Recently, I found the web sitehttp://www.spatial-literacy.org/ (no longer available). This site, developed by Professor Paul Longley at UCL, offers (free) geographical information as to surname profiles (the Surname Profiler) in Great Britain for the years 1998 and 1881. Clearly 1998 is less relevant in origin tracing, as the last century has seen much increased mobility. Interestingly, however, the map derived from their data still shows Most areas in the northern part of the Lake District, Lancashire and somewhat surprisingly Cambridgeshire and Lincolnshire. If you are not familiar with the county boundaries pre the 1974 changes, see http://www.genuki.org.uk/big/Britain.html.
The 1881 version is clearer as to origins, with Most in Westmorland and Lancashire only, and the next category in Norfolk. Next down again was Cumberland, and then Worcestershire, not, as might be expected, Gloucestershire. Various other counties showed some Sauls, including Durham and Dumfriesshire (the map includes Scotland and Wales). At this stage, the evidence seemed reasonably clear. The Sauls originated in the Lake District, with quite separate Saul groups in Norfolk and Worcestershire. There were small groups in other areas, but notably few around London or Yorkshire. This all seemed to fit with what was already known, albeit using maps with a relatively coarse Most to Least marking. Other data available on the site was more detailed, but not as to exact numbers per county. In total, there were 1716 Sauls in 1998, and 1532 in 1881, respectively 43 per million of the population in 1998 and 57 per million in 1881. In 1998, Northern Ireland had 15.5% of the GB rate, the Republic of Ireland had 10.8%, while Australia had 133.5%, New Zealand 92.4%, US 118.6% and Canada 107.1%. These figures are of interest, but should not be over-analysed. One possible inference is the US figure, which, since there have been many other nationalities immigrating to the US, suggests that Sauls may have figured highly as emigrants. The same could be said of Australia, which in recent times has had numbers of non-U.K. immigrants, but remains strongly based on U.K. stock.
I contacted John Slaughter at this stage, and he suggested I write up some of my findings. I agreed to do so, but in thinking about the baseline data, I wondered if an alternative approach would be useful.
The oldest reliable data base which covers the whole country is the 1851 census. The 1841 is apparently not fully available and does have gaps. Accordingly, I chose to look for data from the 1851 census. This is 30 years before the Surname Profiler data from the 1881, i.e. about one generation. Obviously a full census from say 1800 or earlier would have been better again, since less mobility would have occurred; but such information is not readily available. Parish registers could be used, with reservations as to the completeness of the data, but until they are fully transcribed, the work would be impractically large. Another alternative is the IGI; a group of years could be selected, say 10 or more, and the data examined as below. Again, lack of completeness of the data would be an issue, and could easily skew any results. I will look at this at some future date.
Ancestry.co.uk offer a quarterly subscription which makes available all of the census records for England and Wales (separately) from 1851 to 1901. In addition to specific information on individuals, it is possible to determine for example all of the individuals with a given surname in a defined county. I chose to use the 1851 data to determine a county by county spread of the location, at the census date, of all Sauls in England. This was intended to give me an indication of geographical spread, and possibly to confirm or otherwise the evidence to date on the origins of the name. Of course, alternative criteria could have been used, such as place of birth, or selection of heads of household only. This latter has attractions in that some counties are represented by a single family only, while others have a number of small families. I chose to list all individuals, but effectively disregarded the counties with small numbers in making some general comments below.
Total Sauls in 1851
As shown above, there were 1028 Sauls listed in the 1851 census. The total population of England and Wales was 17.9 million at the time of the census; of those, about 16 million were in England, so the rate was ~64, slightly higher than that for 1881. The Great Britain population was 20 million in 1851, up from 8 million in 1801, so earlier data would have been very interesting.
As might have been expected, Cumberland was at the top of the list. If the three counties of Cumberland, Westmorland and Lancashire are added together, 316, they represent 30% of the total. The surprise areas were Middlesex at 142 and Yorkshire at 131, more in each case than Lancashire alone. Grouping Middlesex and Essex with Surrey, purely on geographical grounds, shows a significant cluster in the south-east, not easily identifiable with earlier known concentrations. Perhaps this was just a London effect in the early 1800s and earlier, but it was not evident from the 1881 survey noted above. The Yorkshire numbers are also problematic. Although some were noted in 1881, the 1851 numbers were significant. The most likely explanations are either that they were an offshoot of the Lake District group, or that, if the SJOLFR explanation is accepted, that it may have given rise to more Sauls separately.
Slightly more predictably, Norfolk showed 118, again more than Lancashire, and this supports the Salle village source. Suffolk, with 18, could include some from the same source.
Also surprising was that Gloucestershire was not well represented, at 10. These were primarily (9) in one family, and that a recent immigration from Ireland as noted below. Possibly Saul families had moved to Worcestershire, 30. This is in accord with the 1881 diagram. The relatively small numbers should again not be over-interpreted, since they could stem from a single roving individual only two or three generations before the census. Indeed, the Hampshire group of 4 was a single family in one household. Dorset had none, so the Hampreston group of the Domesday book had presumably died out by 1851. A female Saul, an unmarried 37 year old mine worker, was born in Cornwall, but this was most likely due to a corruption of Sawl(e), see below.
The Oxfordshire Sauls, 36 in 1851, were notably around Banbury, and are supposedly an offshoot of the Saul-Gloucestershire group. The Warwickshire group, 44, is significant in number, and not clearly associated with the others. John Slaughter points to the location, primarily around Leamington Spa, (or Leamington Priors as it was then called). Johns research established that these Sauls can be traced back to two brothers who, almost certainly, came from Horley in Oxfordshire in the early 1800s.
Another county currently known for Sauls is Northamptonshire. Only 2 were recorded in 1851, while there are many more now, including incomers such as myself and my family. Others near me now are from the Leicestershire branch, related distantly to my own family tree; there were none in Leicestershire in 1851. John has pointed out that that the Northamptonshire Sauls are also an offshoot from Horley, particularly the Sauls at Byfield. Horley and Byfield are only a few miles apart though divided by a county boundary. It was the Byfield Sauls who added the extra “L”, and that is why I found on the 1851 census for Northamptonshire 18 Saulls but only 2 Sauls. The evidence suggests that, as far as the Sauls of the South Midland counties are concerned, that Saul and Saull are closely connected. There were Sauls in Horley as far back as the start of the parish registers there in 1547. There is no obvious source for the Saul name in that area, so it has been speculated that the original source is the village of Saul in Gloucestershire and that there had been a migration between the two areas at some time between the development of surnames (perhaps late 14th century) and the early 16th century.
There are of course other possibilities. Of the 1028 Sauls in 1851, 19 gave their birthplace as Ireland. This was the time of large scale emigration from Ireland. We cannot compare this figure directly with the county figures, since these were people who chose to leave their homes and come to England, albeit propelled by famine. Since it was close to 2% of the total, it does have some significance, especially in view of the later low rate for Ireland as a whole. Delving deeper, of the 19, 9 were in one family, including a son born in 1850 in Ireland so they were newly arrived. There were a total of four sons, and they lived in Cirencester, Gloucestershire. Of the 10 Sauls in Gloucestershire, 9 were from Ireland, and by implication identified with Saul in Co. Down. The tenth was a 13 year old female agricultural worker head of her own household. Thus, in 1851, the Saul village Saul surname link was essentially absent.
|Total Saulls in 1851||101|
The Northamptonshire link between Saul and Saull has been commented on above. Clearly from the table above, the Lake District Sauls only used one L, while the Northamptonshire Saulls at that time greatly outnumbered the Sauls. The Middlesex/Essex connection for Saull was also strong.
|Total Sall in 1851||340|
A greater puzzle was for Sall. A partial table is shown above. These are not immediately located to Salle in Norfolk, although some may be. Inexplicably from my work, they were concentrated in Staffordshire. Checking the Birthplace of heads of household also showed predominantly Staffordshire, although there were some immigrants from surrounding counties. There was no obvious family grouping of location. A possible explanation is a family immigration to the area from Norfolk, probably several generations before 1851. John Slaughter mentioned the possibility of transcription errors, from Salt. A check on this showed that it certainly had occurred in some cases, but definitely not in others. Looking at the later censuses, in 1881, the Ancestry transcriptions show 58 Salls, 1891 has 90 and 1901 has 121. John has only one family of Salls in Staffordshire in 1881, and then they were transcription errors. The differences are such that definitive evidence would require examination of each record again in detail, a task I shall pass to a Sall researcher.
Salle is a less common name variant. There was a preponderance in Middlesex, but the numbers pointed only to the possible attractions of London. Main addresses were St. Pancras and Westminster St. James. There were three De La Salles in the census. Two of these were a married couple in Fareham, birthplace Norfolk, and, in St. Pancras again, Alphonse De La Salle, whose birthplace was France. His wife and the rest of the family are shown as Salle in the transcription, but clearly do in the original record. Since there were three sons, it is reasonable to suppose that they had successors; by 1901 there were just two De La Salles, a widow and her son. It is interesting to note that, as a surname, De La Salle may have entered the country as late as the 1800s, in addition to the Norman Knight. A search of the current BT online directory gave a single De La Salle, when looking for Salle, as the De La is not recognised.
Finally, Sawle is very clearly from Cornwall. In the 1881 profiler, it was apparently the only location, i.e. almost all occurances were in Cornwall. Repeating the above technique using Sawle and the 1851 census showed:-
There were just four Sawls without the e. Of these, two, in Camberwell, were clearly transcription errors for Saul. The third, in Lancashire, was a lodger and was most likely another Saul, so these three should be added to the Saul grouping. The fourth, in Cornwall, had a wife with the spelling Sawle, so again a transcription error was obvious. This completely eliminated the no e variant. By 1901 there were 10 Sawls, but careful examination almost eliminated these too. Four, in one family, could credibly have been Sauls, with transcription errors. This page had then been exactly duplicated in the record (page 25) making 8. The other two could have been the result of misspelling or misrecording, but did appear as Sawl.
In 1901, the Sawles were well dispersed through the southern counties, although centred on Cornwall. There is no Sawle on current maps, but Sawley exists both in Lancashire and Yorkshire. This is likely to be Sawls ley, i.e. Sawls meadow, or of course Sauls meadow. However, I cannot reasonably associate this with the Cornwall grouping.
Some years ago, the TV programme Meet the Ancestors did an interesting DNA profile of the entire country. Of course, after perhaps 200 years of significant population movements, many issues were clouded, but the one clear statement was that the population of the Lake District contained a significant proportion of Norse, i.e. Norwegian input. This fits well with the Norse theory of SJOLFR. Of course, Yorkshire was a stronghold of the Danes, i.e. Norwegians, Danes and Swedish Vikings for about two centuries, so the same genetic input would apply. Similarly, the village of Saul is located exactly where raiders going up the Severn estuary would land, especially the predominantly Norwegian raiders of the west coast. Further groups would have gone around the coast to Hampreston, again with easy access from the sea. The very low scores for the south-western counties possibly indicate relatively little Norse influence, although at some point virtually the whole country was invaded by them.
The Salle connection may be more difficult to isolate genetically, since only the De La Salle family would have had Norman DNA, and it is unlikely that this could be separated from the Norse; the Normans were mainly Vikings who had lived in Normandy for 150 years.
So, what does that all mean? Although the 1851 data is reasonably sound, there are some caveats. I have noted above some of the difficulties posed by transcription errors and in one case a duplicated page (1901 census). There could be more errors of this type. It is known that much data from the Manchester area 1851 census has been damaged or destroyed.
Some years ago, living in Manchester, I noted that the local phone book had more Sauls than any other in the country outside London. The 1851 figures show no Sauls at all, presumably due to the destruction noted above. In 1861, there were 9 Sauls in Manchester, including two heads of household. Of these, one was born in Kendal, with other family members born in Preston, suggesting a shift in that generation through Lancashire to Manchester. The other was born in Cheshire. In total, 6 would have been alive in 1851 and possibly, but not definitely, living in Manchester.
The picture of three main English sources, the Lake District, Norfolk and Gloucestershire Worcestershire is largely confirmed. The Banbury and Warwickshire groups are from the Saul/Gloucestershire source. The large Yorkshire numbers were some surprise, speculatively originating from the Lake District group, but possibly independent. The Middlesex connection should be added, although the exact reason for the numbers and their origins is not clear. It may be just the gravitation to London, possibly from all of the other groups, which has happened over many centuries. The connection with Ireland was also noted, and dominated the 1851 Gloucestershire figures due to a single family.
As a little aside from the main work, I looked at Ancestry.co.uk for New York immigration. Between 1846 and 1851, there were 35 Sauls immigrating to the U.S., respectively 11 from Great Britain and 24 from Ireland. One ship was from Greenock, while the others were from Liverpool. This tends to confirm that Liverpool Sauls may have more than one source; not all arrivals from Ireland completed the crossing to the U.S. Two who did were both listed with the age 800, perhaps the most obvious error I have seen so far.
A DNA analysis of Sauls would be a most interesting exercise, but would be especially complex in the interpretation.
I have to thank John Slaughter for a number of useful comments on the draft of this article. John was able to clear up several points which I could not, and he has made a valuable contribution to this (and many other!) articles on this topic. Also at Johns request I have added the Salle, De La Salle and Sawle variants. If any member wishes to use similar techniques, either for different Saul analyses, or for the other surnames in the Society, I will be pleased to advise.
 The Liverpool Sauls almost certainly have multiple sources, including distant parts of my own family tree.
Anglo-Saxonized spellings for the word were “Soelf” and “Seulf”, leading to the rather uncommon surname Self.
 Cameron, K. (1988; 1996) English Place-Names, London: B.T. Batsford Ltd.
Loyn, H. R. (1962)Anglo-Saxon England and the Norman Conquest, London: Longmans, Green and Co Ltd.
 I have used the term Lake District as a collective term for the area including Westmorland, Cumberland and that part of Lancashire located there. Broadly, it is modern Cumbria, but I never liked that term, and as a Lancastrian, I want that part of Lancashire returned to where it belongs.
 Lancashire here is the traditional Lancashire, including roughly what is now Lancashire and Lancashire over the water, the detached section in the southern part of the Lake District. Almost all Sauls in Lancashire are descended from the Lake District group, but in some cases branching off before 1600.