Drive through Lincolnshire and you pass Grimsby, Whitby, Scunthorpe and dozens of other places whose names sound Old Norse. The same goes for Caudebec and Yvetot in Normandy. Such endings are the most visible everyday legacy of the Viking Age. Yet they prove less than people like to think: a place-name tells you which language someone spoke when the name arose, not how many people migrated or where exactly they came from. This article explains the main building blocks, tests the famous examples and names the traps.
Place-names are tenacious. They outlast conquests, dynasties and changes of religion, because the people on the spot simply keep using them. That is exactly why they matter for the Viking Age: where chronicles are silent, or report from far away and decades later, the name of a village still stands there and preserves a piece of the language spoken when it was coined.
At the same time a place-name is a difficult witness. It carries no date. It does not say who gave it, or for whom. And it becomes visible in writing only centuries after it arose, often in the spelling of a Norman scribe who understood Old Norse or Old English only half. Anyone who wants to recognise "Viking place-names" has to do two things at once: know the building blocks and know the limits of the method.
This page deals with Norse settlement names in England, Normandy, Scotland and Ireland. Whether Odin, Thor or Freyr are hidden in place-names is a separate topic, covered in Gods in Place-Names. Here we deal with the ordinary: farms, villages, streams, clearings.
Most English place-names have two parts: a first element, often a personal name, and a second element that says what kind of settlement it is. For the Norse names in the area of the Danelaw, overview pages such as viking.no list the following second elements. Their explanations are brief and serve as rough orientation, not as a dictionary.
The last item already shows the basic problem. Old English and Old Norse were closely related. An ending that both languages knew is not unambiguous. The overview mentioned above states explicitly that endings such as -borough or -bury were used by Anglo-Saxons and Scandinavians alike.
The first part is often more reliable than the ending. If a place-name contains a Norse male name (in Grimsby, the name Grímr), that points to Norse naming. The name scholar Gillian Fellows-Jensen re-examined the situation in East Anglia in a paper in the journal Nomina (volume 37). Her result is tellingly sober: of 27 East Anglian -by names she looked at, only eight certainly contain a Norse personal name. For Herringby she revises her earlier reading: it probably contains the Old English word for "herring", not a Norse name. For Thrigby, too, she calls her interpretation "by no means certain".
That is the lesson in miniature: even a specialist who has known the material for decades leaves part of the readings open and corrects them over time. A holiday-maker who spots a signpost and shouts "Norse!" is often right, but not always.
Grimsby on the Humber is the textbook case of a wholly Norse name. The ending -by is Old Norse, the first element a personal name, in the usual reading Grímr. According to the English Wikipedia article, Domesday Book of 1086 is the earliest dated record of the place, listing about 200 people, a priest, a mill and a ferry. I have not checked that against the Domesday edition itself.
A founding story goes with the name: the fisherman Grim is said to have founded the town. It comes from the medieval verse romance about Havelok the Dane, which was composed much later than the settlement. Historians treat the tale as a legend, not as evidence for a real founder. That a man called Grim existed cannot be proved; it is equally possible that the name led to the story, not the other way round.
Another reading links "Grim" to Odin, who hides as Grímnir, "the masked one", in the Edda. Places with Grim- in the Danelaw and beyond are partly explained this way. For Grimsby this is called likely, but it is an inference, not a secure finding; the names cannot be resolved with certainty in one direction or the other. When exactly Danes settled Grimsby is likewise answered only roughly by the sources: the 9th century is named.
Whitby in North Yorkshire shows how a name changed without the place disappearing. In Anglo-Saxon times it was called Streoneshalh (Streanæshalch in Bede); its meaning is uncertain, according to Wikipedia "Streon's nook of land". According to Wikipedia the modern name first appears in Domesday Book and is read there as "Hvíta's farmstead", from Old Norse Hvítabýr. The reading is not unanimous, though: other accounts take hvítr ("white") plus býr, i.e. "white farm" or "white settlement", and admit that nobody is sure what the "white" refers to. So the Norse name prevailed over the older English one, an indication of conditions in which Norse speakers on the spot set the tone, nothing more. The source says nothing about a population replacement, and anyone who reads one out of the name overstretches the evidence. A chronicle passage shows what land-taking looks like in the written record: for the year 876 the Anglo-Saxon Chronicle reports of Healfden (Halfdan), a leader of the Great Army:
"The same year Healfden divided the land of the Northumbrians; so that they became afterwards their harrowers and plowers."Anglo-Saxon Chronicle, entry for the year 876, English translation (translator not named on the source page; older, public-domain text), wording seen via the retrieval tool in an online edition (Avalon Project)
Scunthorpe in Lincolnshire is a steel town today and was a hamlet in the Middle Ages. According to the English Wikipedia article, Domesday Book records it as Escumesthorpe; the reading leads to Old Norse Skumasþorp, "Skuma's homestead", with Skuma as a personal name. The ending -thorpe fits the sense of "dependent secondary settlement": an outlying farm laid out or worked from a larger place.
What happened later is interesting. The town's coat of arms shows five villages that grew together: Crosby, Scunthorpe, Frodingham, Brumby and Ashby. In 1919 the urban districts of Scunthorpe, Brumby and Frodingham were merged with the parishes of Crosby and Ashby. Crosby and Brumby both end in -by; I have not analysed each of the five names here and offer that as an observation, not as proof. A landscape with villages in -by and -thorpe side by side is in any case no isolated case.
The ending -thwaite goes back to Old Norse þveit, "clearing". That the same word occurs in Normandy is clear from the French forms: places such as Le Thuit or Bracquetuit are traced to þveit in the Wikipedia article on Norman toponymy. In England's Lake District you meet the word in many small places in -thwaite. Overview pages link it rather to areas where Norwegians settled. I have not checked that attribution against the specialist literature; research discusses it in more differentiated terms than an overview reports.
Kirkby is a widespread type among the church names: Old Norse kirkja, "church", plus býr, "village": the village with the church. According to the Wikipedia article on the Danelaw there is a "remarkable number" of such names, and at some sites remains of Anglo-Saxon church buildings were found (the article cites Taylor, Anglo-Saxon Architecture, 1965). That is a fine clue: the name was formed in Norse, but the church was already there. A community had given its older church a new, Norse name.
Scotland offers a related example: names of the type "Kirk- plus saint's name", such as Kirkoswald, point according to one overview to a mixed Gaelic-Norse population, though it admits that this interpretation is "not certain".
Besides purely Norse names such as Grimsby there are mixed forms: a Norse personal name plus an English second element such as -ton (Old English tūn, "farm, village"). The model case is Grimston, which gave the research its term "Grimston hybrids". Fellows-Jensen counts over eighty such names in East Anglia, spread across the region except the Fens and Flegg. Where -by names are dense, the hybrids are rare; in her paper this observation goes back to Insley.
What does that mean? Mixed names indicate that Norse names reached into settlements named in English. Their bearers could be wealthy, but also hold tiny plots, she writes. And the names may be younger than the settlement itself; many small holdings probably arose later. So whoever sees a Grimston hybrid is not looking at the landing place of an army but at a society in which English and Norse naming had long interlocked. How the settlement was embedded politically is described in the article on the Danelaw Danes and the Five Boroughs.
The viking.no overview speaks of "many hundreds" of Scandinavian or hybrid place-names in the Danelaw and names Lincolnshire as the area with the greatest density of -by names. It gives no exact total, so I give none. In East Anglia, 13 of the 27 -by names Fellows-Jensen considered cluster in Flegg. As possible explanations she offers: Flegg as a Viking base in the 9th century, as a Danish enclave after 917, or as a strategic point protecting trade. Which of these is right is open.
Density alone, then, is no answer to "how many". It first asks: who distributed land here, who gave names, who was asked when the names were written down?
Across the Channel the evidence is just as tangible. The region that later became Normandy came under the rule of Norse leaders in the early 10th century; how that happened is told in the article on Rollo. The English Wikipedia article on Norman toponymy lists as the main types:
The traps are interesting. For the ending -beuf the article warns that later spellings such as -bœuf ("ox") rest on folk etymology; some places really do mean oxen, others do not. Two names, Criquebeuf-la-Campagne and Carquebut, are doubtful and might come from Old Norse bú. Honfleur, Barfleur and Harfleur look Norse, but the article derives -fleur from flōd or flēot ("run of water, river going into the sea") and does not list it among the Norse types; whether Norse or Old English is thus not settled. It traces -londe to Old Norman londe ("forest"); I have not checked other scholars' readings. Many of these readings are qualified in the article with "probably" and "most likely", because Danish, Old English and Norse forms resembled each other.
On the origin of the settlers it says that most probably came from the area of present-day Denmark, less often from Norway; after the founding of the duchy more arrived from the Danelaw, some from the Scottish islands and Ireland, especially into the Cotentin. Norse speakers were absorbed linguistically within a few generations. This summarises scholarly opinion that I have not checked against the works cited. The main authorities the article names are Adigard des Gautries (1954) and Renaud. For the later Normans see also Vikings, Normans and Varangians.
The picture in Scotland differs greatly by region. In the Northern Isles, according to the Wikipedia article on Scandinavian Scotland, place-names of Celtic origin are "few" in Orkney, and Pictish name elements are "virtually non-existent" in Shetland (exceptions: Fetlar, Unst and Yell). In the Outer Hebrides, Coll, Tiree and Islay, pre-Norse names were almost completely obliterated. In Caithness and Sutherland the Norse character persisted into the 13th century. South of Beauly the article finds no Norse place-names on the northern mainland. On the west coast, Norse names often appear alongside existing settlement rather than in place of it. In Galloway and Dumfries the situation is "complex" and of mixed Gaelic, Norse and Danish influence (Crawford, 1987, according to that article).
The caveat the article itself makes matters: the place-names date from a relatively late period, so the timing of the shift from Pictish to Norse remains uncertain. And the spread of the Norse language in the Hebrides is inferred rather than directly attested. For more, see Vikings in Scotland, The Earls of Orkney, Jarlshof in Shetland and The Isle of Man.
In Ireland it is mainly harbour towns: according to the Wikipedia article on the Norse-Gaels, Norse settlers founded independent kingdoms at Dublin, Waterford, Wexford, Cork and Limerick, which did not survive the Norman invasions but whose towns kept growing. For individual Irish place-names I cannot offer checked etymologies here because I have not verified them in specialist literature; so I name none. One example with a caveat: Oxmantown in Dublin is said to go back to "Austmanna-tún", "homestead of the Eastmen", and the source itself marks this as unproven. More in Vikings in Ireland and Viking Dublin.
The central question: what can the density of Norse names tell us about the number of immigrants? The short answer: very little. The viking.no overview states explicitly that a Scandinavian name does not prove Scandinavian settlers: naming customs spread among English speakers, names were sometimes given by outsiders, and some arose decades or centuries after the Scandinavian population had assimilated. The Wikipedia article on the Danelaw notes that Old East Norse and Old English were partly mutually comprehensible and that contact brought many Norse words into English (it cites Biddulph, 2003).
Place-names thus show first of all that on this land, at a certain time, Norse was spoken or Norse names were fashionable. A headcount cannot be derived. A village may have received its Norse name from a small elite that allocated land; and conversely there may have been many newcomers whose traces did not stick in names.
Modern genetics adds to the picture but does not replace name studies. According to the Wikipedia article on the Danelaw, a 2020 study found evidence of a major influx of Danish settlers in the Viking Age, while a BBC-commissioned survey of 2000 was not designed to tell descendants of Danes from those of Anglo-Saxons, because both groups came from overlapping North Sea coasts. I have not read the study myself. How hard such data are to interpret is shown in Viking Genes: What the Study of 442 Genomes Really Shows and Am I a Descendant of the Vikings?.
First: ending equals origin. -wick, -borough and -bury are not unambiguous. Even -by can in single cases be late and formed in an English setting. The test is the earliest spelling: how was the name written before 1200, and what is in the first element?
Second: folk etymology. Caldebec becomes Caudebec, -beuf seemingly becomes "ox". The modern spelling is often a poor guide.
Third: personal name without a witness. The fisherman Grim, Skuma, Hvíta: these are men inferred from names. Nothing is known of any of their lives. In the case of Grimsby the legend speaks of a fisherman, the research only of the name.
Fourth: dating. A name may have arisen centuries before its first written mention, but also only after the settlement. Early written attestation often lies in Domesday Book, thus at the end of the Viking Age. Whether the name is older cannot be read from the record.
Fifth: confusion with god-names. Not every Grim- means Odin, not every Thors- Thor. That belongs in another article; see Gods in Place-Names.
A workable procedure: first find the oldest spelling, in a county place-name volume or in English name studies, not in a travel guide. Second, identify the second element and check whether it existed in both languages. Third, look at the first element: personal name, landscape word, church? Fourth, ask whether there is a parallel across the North Sea, such as Caldbeck and Caudebec, or Kirkby and Kirkeby. Fifth, include the neighbourhood: is the place in an area with dozens of similar names or alone? And sixth, never confuse the result with population numbers.
Even specialists reach different results for single names. Where I write "according to" in this article, I saw the reading only through the summary named, not in the original literature.
How do I recognise a Norse place-name in England? By second elements such as -by, -thorpe, -thwaite, -toft or Kirk-, especially in Lincolnshire, Yorkshire and the north-west. Certainty comes only with the earliest spelling and the first element.
What does Grimsby mean? In the usual reading "Grim's village": Norse personal name plus býr. The story of the fisherman Grim comes from a medieval romance and counts as legend.
Do place-names prove how many Vikings came? No. They show that Norse language and names played a role, not the number of immigrants. Naming fashions could spread among English speakers.
Are there Norse place-names in France? Yes, mainly in Normandy: -tot, -bec, -beuf and -thuit. Some readings are uncertain because Danish, Norse and Old English forms resemble each other.
Does Whitby mean "white village"? That is uncertain. The English Wikipedia article reads "Hvíta's farmstead", after a personal name; a visitor guide reads hvítr, "white", plus býr. I could not settle which is right from a place-name dictionary.
Gods in Place-Names · The Danelaw Danes and the Five Boroughs · Vikings in England · Jorvik · Guthrum · Rollo · Vikings, Normans and Varangians · Vikings in Scotland · Vikings in Ireland · Viking Genes
Gods in Place Names · The Danelaw · The Vikings in England – from Lindisfarne to 1066 · The Norse of Jórvík · Vikings in Scotland.
All History & Finds articles →
Oslo celebrated its 900th birthday in 1950 and its 1,000th in 2000. Count backwards and the two founding dates sit half …
Read more →When we picture Vikings, we usually see longships off English monastery coasts and sails over the North Sea. But an …
Read more →If you know the Viking Age only as a string of raids on monastery coasts, you have not seen the Baltic. Between …
Read more →On screen, nearly every Viking wears leather, studs and fur. In the museum, the entire age offers one nearly complete …
Read more →