"Yorkshire, particularly the East Riding, was subjected to very heavy Scandinavian settlement which may partly account for the paucity of Celtic names" (Faull, 1977: 15)

In the following text, the terms "Celt" and "British" are employed interchangeably. References to "Anglo-Saxon" and "English" are similarly equivalent.

Barber (1993) notes that the greatest linguistic influence of the Celts was in terms of place-names. Notwithstanding, Celtic place-names are far exceeded by those of English origin and the tendency is that Celtic survivals are more common in areas of late penetration by the Anglo-Saxons. This assertion is examined in greater detail below.

Starting with an overview of the nature of Celtic name-giving, the correlation between the survival of British river-names and the English advance westwards across the country is then examined. Consideration is then given to the subsequent adoption of Celtic place-names by the incoming Anglo-Saxons. Finally, the focus is narrowed to Yorkshire and the Celtic place-names still extant within the county.



Faull (1977) suggests that the endurance of British place-names in the Anglo-Saxon period is not necessarily indicative of Celtic settlement. It may simply be that the English adopted the names from nearby British speakers. In cases other than places of particular importance (such as those containing the elements duno- ("fort") and duro- ("walled town") relatively few earlier Celtic place-names are indicative of Celtic settlement. Unlike the Anglo-Saxons who employed precise terms to specifically identify the type of habitation (e.g. homestead, farm etc), British place-names are predominantly topographical rather than habitative, i.e. they named their settlements in relation to neighbouring topographical elements. (e.g. Crayke (NR) "rock, cliff" and Roose (ER) either "heath", "moor" or "promontory" "hill").

Faull proposes that either Celtic settlements were so similar to each other that there was no reason to distinguish them in terms of size in the manner of the Anglo-Saxons (e.g. ham, tun, wic, wor) or that naming settlements after the surrounding topographical features was merely a characteristic of Celtic name-giving.


Celtic place-names

According to Faull (1977) Celtic place-names may be broken down into three categories:

The first two groups generally tend to be Old English formations and they are accordingly discussed in the section relating to English place-names. With regard to the final category, Gelling (1988) adds that, in relation to Celtic and pre-Celtic place-names, two general principles can be said to exist:

It is notable that, in areas where Celtic place-names are rare, only the larger geographical features bear Celtic names. Conversely, in locations where such names are frequent even minor sites and features bear British designations.

Faull (1977) explains this phenomena by drawing attention to the correlation between the progress of the Anglo-Saxons from east to west and the survival of Celtic river-names. She cites Jackson (1953) who has demonstrated that the frequency of such names shows a marked correspondence to the westward movements of the invaders.


British River Names

As can be seen from the Map 1, Jackson divides England into four areas running from east to west.


Map 1

After Cameron 1996: 46


Area 1
This area covers everything east of a line running from the Yorkshire Wolds down to Salisbury and the New Forest. It correlates with the territory of primary English settlement around 600 AD and contains very few British names. Those that do exist are generally the large or medium rivers and a high proportion of these are of doubtful etymology, possibly being of pre-Celtic origin.

Area 2
This is the area of intermediate settlement and reflects the English penetration westwards during the during the early 7th century. It runs to the east of Cumberland and Westmorland in the north to the Ribble and thence from Chester to the Bristol Channel along the Dee and Severn valley. Skirting Somerset it passes Selwood and onwards to the Hants-Dorset border. A much larger number of British river names exists in this region added to which their etymologies are more certain. Furthermore, a number of small rivers bear Celtic names.

Area 3
This area correlates to the final stages of the English advance westwards. Located to the west, it includes Cumberland and Westmorland, west Lancashire, the Welsh border counties and south-west England as far as the Tamar. The proportion of river names in this area is high and Celtic etymologies are more certain. There are more woods and hill names than in areas to the east and even stream names are of Celtic origin.

Area 4
This area, which covers the territories of Cornwall, Wales, and the south-west corner of Herefordshire, falls outside the limits of the English territorial settlement. The nomenclature within is almost purely Celtic.

The question arises as to why the Anglo-Saxons adopted some Celtic place-names and not others.


Adoption of Celtic Place-Names

Just as pre-Celtic names would be passed on or taken over by the Britons from the pre-existing Bronze Age culture, so Celtic names were passed on to the Anglo-Saxons. However, this did not happen all at the same time because, as noted above, different parts of England were occupied at different times. The spread of the Anglo-Saxons beginning with the penetration into Kent and coastal areas of the East and Southeast from the middle of the 5th century lasted into the 9th century when advances were made into Cornwall. The very fact that names were adopted by the English presupposes the survival of Celtic speaking people in the districts where the names occur.

Cameron (1996) states that, although it is uncertain how these names were communicated, one suggestion is that their adoption by the Anglo-Saxons was the result of bilingualism, specifically among the Britons. In other words, the place-names were passed on rather than borrowed. Notwithstanding, it would appear that, in those territories falling under their immediate control and occupation, the English either failed to adopt Celtic place-names, or changed them to an unrecognizable form. They did however adopt British names for locations on the fringes of these areas and in areas of later settlement.

Faull (1977) proposes two possible reasons why the English failed to absorb local British names place-names in areas of early settlement. Firstly, because they were wholly occupied with establishing a foothold in the new territory, they were either disinclined, or had insufficient time, to form any sort of close association with the indigenous population. However, as the English had in fact established some settlements during the period of Roman occupation resulting in the English and the Romano-British living in "close proximity", Faull considers this explanation to be somewhat questionable.

She considers that a much more reasonable explanation is that the lack of bilingualism existed within the two groups. As such, the transfer of place-names from one community to the other was impeded. Alternatively, it is possible that the English had already named their settlements and the neighbouring topographical features before the development of bilingualism within the locality. This would explain why British names tend to occur on the fringes of the primary English settlement areas rather than centrally.


Celtic place-names in Yorkshire

In relation to Yorkshire, the East Riding is situated in Jackson's Area 1 whilst the remainder of the county falls within Area 2. Concerning the first phase of the English incursion into the county, although there is a dearth of British names on the North Yorkshire coast, the Wolds and the Vale of Pickering, there are some on the "coastal strip of good land" and on the fringes of the Wolds (see Map 2).



Similarly, British names are rare in Wensleydale, the Howardian Hills and the Vale of York, areas which were penetrated by the Anglo-Saxons in the second phase of their expansion. Those that do exist occur in the margins of these areas.

The third and final stage of Anglian push westwards occurred in the early 7th century when they advanced into the West Riding. A large number of British place-names exist around the region of Elmet probably due to the fact that this area maintained its independence until the late 7th century.

Further details of the walh, cumbra, brettas etc place-names may be accessed by clicking here.

Citing Jackson (1953), Faull claims that, because of the probable changes arising out of phonological substitutions and "folk etymologies", scarcely any pre-English names of pre-1086 vintage exist in their original form in Yorkshire.

The following etymologies have been drawn from several sources including Cameron (1996), Gelling (1988), Leith (1996), Strang (1970), Thomson (1964) and Thurlow (1979). The presence of an asterisk before a word (e.g. *penno) is indicative of a postulated or constructed form. Any reference to Old English is abbreviated to OE.


River names


Possible or probable etymology
Aire (WR)
Possibly derived from British *Isara "strong river"
Calder (WR)

Possibly from a word meaning "violent" and hence "the rough water".

Dearne (WR)

This may be of Celtic origin but no acceptable etymology has been suggested.

Derwent (ER)
From British *Derva "oak" and thus probably "river lined with oaks".
Don (WR)
Probably from British *dana from the root *dan (cf the Danube and the Don in Eastern Europe)
Dove (WR)
From the British dubo-"black" hence the "dark river".
Esk (NR)
Derived from British Isca (cf Romano-British Isca as in Exe. Exe is an Old English metathesis of Esk).
Hodder (WR) Suggested meaning is "peaceful" although research has shown that Celtic etymology is doubtful. In all likelihood the name is pre-Celtic.
Hull (ER)
Originally regarded as Celtic but etymological reasoning is flimsy and the reconstruction debatable.
Humber (ER) The suggested Celtic derivation is translated as "good" or "health giving" but this etymology is regarded as doubtful on the same grounds as for the River Hull. It is possible that the name is in fact pre-Celtic.
Kyle (NR)
From Celtic word meaning "narrow".
Laver (WR)
From labara "talkative, noisy" and thus"murmuring river".

Lune (WR)

There are two lines of thought. Either the name is derived from British lano "full" or from an alternative word meaning "good" or "health giving".
Nidd (WR)
Is of uncertain etymology but is possibly from a root meaning "brilliant.

Ouse (ER)

Many texts books suggest Celtic origins (possibly from the root *udso) but the proposed etymology is unconvincing. The name may therefore be pre-Celtic.
Rother (WR)
Probably means "chief river" from the Celtic prefix ro- and dubro "water".
Rye (NR)
Possibly from a word cognate with Latin rivus "stream" (Also the source of Riccall meaning "calf of" or "little Rye").
Tame (NWR) From a Celtic adjective meaning "dark"
Tees (ER)
Possibly from a word meaning "surging".

Ure (WR)

May be based on the river-name *Isura linked to the Roman station of Isurium at Aldborough. Following the loss of intervocalic /s / and under the influence of the Old English diphthong io/eo the outcome is Eor. This subsequently becomes Scandinavianised to Jr from whence the modern form of Ure.
Wharfe (WR)
Suggested meaning is "winding river".




Possible or probable etymology
Craike Hill (ER)
The origin is identical to that of Crayke (NR)
Penhill (NR)

This is a tautology meaning "hill-hill" (from British *penno + OE hyll.). *Penno as proper noun was taken by the English to be a proper name.

The Chevin (nr Otley, WR )
Derived from Old Welsh cefn "ridge".


Areas / districts


Possible or probable etymology


The more northern of the two Anglian kingdoms in Northumbria, the name is derived from British Briganticia (from the tribal name Brigantes).


Names such as Barwick in Elmet and Sherburn in Elmetare suggest that Elmet was in fact an area rather than a place. Generally regarded as a British kingdom it maintained its independence until the late 7th century AD. The origins of the name are uncertain.


Probably a British enclave similar to Elmet. The name is possibly derived from British kramo- "garlic" although the reason for this is unclear (cf Italian Cremona).


The southern Anglian kingdom of Northumbria, the name probably has its origins in British *dobriu the possible meaning of which is "land where there are many rivers".


Towns /villages / cities

Name Possible or probable etymology
Alne (NR) Certainly pre-English and possibly Celtic, it has been suggested that Alne on the River Kyle was originally as the name of the river itself (compare the rivers Allen, Alne and the Welsh Alun).
Beverley (ER) The first element may be from British bebro- "beaver".
Cammock (Settle, WR) Probably originates from a stream-name meaning "bent, twisting".
Catterick (NR) Probably derived from the old name for the Swale.
Craddock (WR) Probably an old stream-name representing the Brythonic form of the personal name Caractacus.
Crayke (NR) From British *krakjo "rock" (cf crag)
Dacre (Ripon) (WR) A stream-name possibly from British dakru "tear", and thus perhaps "trickling".
Dent (WR) A British name possibly cognate of Old Irish dinn "hill".
Ecclesfield (WR) All are British / English hybrids incorporating the British egles (the Brythonic form of Latin ecclesia. The Brythonic form was subsequently adopted into Old English as ecles). Note: Exley =*Ecclesley
Eccleshall (WR)
Eccleshill (WR)
Exley (WR)
Glaisdale (NR) Probably Celtic Glas "blue, grey, green".
Ilkley (WR) Any Celtic etymology is doubtful and the name may in fact be pre-Celtic.
Leeds (WR) Formerly the name of a district rather than a place. Views on its etymology differ but it may have been either the original name for the River Aire or referred to the local populace whose name contained the element Lat- (cf Romano-British latenses). Smith (1961) supports the notion of it origins as a river name with subsequent progression to a folk name and ultimately to the name of a district.
Leeming (NR) Probably a river-name turned settlement-name. The etymology may be linked with British *lemanio "elm-tree". However, other possibilities exist.
Leven (in Holderness) (ER) Probably from a river name which could possibly have been Prolemy's Libnios meaning "smooth" or "gliding".
Penistone (WR) Comprising British *Penno "hill" + OE ing + tun
Rossington (WR) The first element is possibly derived Medieval Welsh ros "moor" (see also Roose (ER) mentioned earlier). However an alternative proposal is the word was not a place-name but a common noun adopted by the English in the same way as they appropriated egles.
York The name originates from either Celtic Caer Ebruac or Eburacon- based on either personal name Eburos or the noun eburos "yew tree"; The Celtic name was later adopted by the Romans as Eboracum. Thence it became Old English Eorforwic and subsequently Scandinavian Iorvik before assuming its present form in the 13th century.
Caer signifies a fortified place (cf Carlisle)



Barber, C (1993) The English Language - A Historical Introduction, Cambridge: CUP

Cameron, K (1996) English place-names, London: Batsford

Faull, M L (1977) British Survival in Anglo-Saxon Northumbria. In Lloyd Laing (Ed)(1977) Studies in Celtic survival, British Archaeological Report 37

Gelling, M (1988) Signposts to the Past, Second Edition, Chichester: Phillimore.

Jackson, K. (1953) Language and History in Early Britain, (Edinburgh, 1953)

Leith, Dick (1996) The origins of English. In Graddol, D., Leith, D and Swann, J. (1996) English history, diversity and change, New York: Routledge.

Smith, A H, (1961) The place-names of the West Riding of Yorkshire, Cambridge, Cambridge University Press,

Strang, B M H (1970) A History of English, London: Methuen.

Thomson, R L (1964) Celtic Place-names in Yorkshire, In Transactions of the Yorkshire Dialect Society Vol. XI, Part LXIV

Thurlow, W. (1979) Yorkshire Place-names, Clapham, North Yorkshire: Dalesman Pubishing.