"One of the most surprising facts about the history of English is the resistance of the language to Celtic loans, though the communities have adjoined and overlapped, and their members have intermarried, throughout the period when English has been spoken in this country - and even before." (Strang 1970: 93-4)

This page opens with a brief overview of the ancient Celtic languages after which the question of borrowings from these languages is considered in terms of both Standard English and dialectic forms. The section concludes with an examination of sheep scoring numerals and their suggested origins.


The Celtic Languages

Celtic, the language of the inhabitants of Britain and the first about which there exists a definite knowledge, belongs to the Western Branch of the Indo-European languages and existed in three forms: Gaulish, Brythonic and Goidelic.

The main feature which distinguishes Brythonic from Goidelic is the development of the Indo-European /kw/. This becomes /p/ in former and /c/ in latter. This situation continues to the present day as can be seen from the following examples.

Welsh Irish Gaelic Meaning
pen cenn head
pair coire cauldron
pedwar ceathair four

Hence the the Brythonic Celts are also known as the P-Celts and the Goidelic Celts as the Q-Celts


Celtic loan words

In relation to borrowings from the Celtic languages, Stalmaszczyk (1997) suggests that Celtic words taken in to English may be grouped into three categories

Category Example
1. Continental borrowings by the Angles, Saxons and Jutes prior to their invasion of England

OE rice from Celtic rix (cf Old Irish ri genitive form rig "king"). This word survives in present-day English as the final element of "bishopric".
2. Brythonic words taken into English from the mid-5th century AD

OE brocc from Brythonic *broccos (cf Old Irish brocc; Welsh broch and Gaelic broc "badger). Still survives in English dialect.
3. Words of Old Irish origin, associated with the church and religion, brought in by the Irish missionaries in the 7th century.

OE dry from Old Irish drui (plural druid) "druid, magician".

After Stalmaszczyk, 1997: 78

Commenting on the above Stalmaszczyk observes that only the Brythonic loans might be considered to be genuine Celtic borrowings. He goes on to observe that the examination of Celtic words in English raises three principle questions:

1. When was the item adopted?

2. How many such words were borrowed?

3. What is the linguistic source of the borrowing ?

In relation to questions 1 and 2, the following table shows the estimated number of borrowings and the approximate period during which they occurred.

Linguistic Period (approximate dates) Number
Old English (450 - 1100) 9 - 23
Middle English (1100 - 1500) 50 - 80
Early Modern English (1500 - 1700) 60 - 85
Late Modern English (1700 onwards) 55 - 102

After Stalmaszczyk, 1997: 78

With regard to question 3, Stalmaszczyk cites Serjeantson (1935) in support of the view that, before the latter half of the 16th century, there were very few words of either Irish, Welsh or Gaelic origin present in Middle English. Nevertheless, as time progressed, more such words were absorbed into the language. However, it is estimated that the final total of such adoptions amounts to no more than 300. The following are some of examples.

Origin Example Etymology


bean "woman" + sidhe "fairy hill"
go "to" + leór "sufficiency"
Scottish Gaelic bog
bogach "soft"
sluagh "army" + gairm "shout, cry" i.e. war or battle cry
Welsh corgi
cor "dwarf" + ci "dog"
gwlan "wool"

After Stalmaszczyk, 1997: 79

In terms of Brythonic loan words, however, few were adopted by the Anglo-Saxons. Other than river names and place-name elements such as comb (valley)and torr (rocky peak), only twenty or so words taken into Old English can be reasonably considered to have Brythonic etymologies. Furthermore, not all the such borrowings survived and those that did did so only in dialect. Brock, which appears in the dialects of all northern counties of England, has already been mentioned. Other examples are:

Celtic Meaning Dialect use Comment
binn basket, crib corn bin OED claims OE etymology and makes no mention of Celtic. Sometimes binge in Yorkshire and Lincolnshire although this originates from ON bingr "heap" (cf Danish bing "bin"). Bing used dialectally for "bin" since the 15th century.
dun dark coloured a dull, dingy, or greyish brown colour Employed in various dialects of Scotland, Ireland and England. OED claims origin is OE dun perhaps from Celtic; (cf Gaelic donn "brown").
bratt cloak pinafore, apron, Appears only in Northumbrian texts, e.g.. the Lindisfarne Gospels (cf Welsh brat "pinafore, rag"). Recorded in SED V 11 2.
carr rock
Appears only in Northumbrian texts.


Sheep-scoring numerals

Early in the last century it was considered by some that this method of counting was of Celtic origin. For example, writing in 1927, J.R. Witty claimed that the scores had been in use in Cumberland, the Pennines and Yorkshire for around fifteen hundred years. He provides examples of several varieties according to geographical location. For example:

Number North Riding (not specified) Middleton - in- Teesdale Craven and N.W. Moorlands
ena yan arn
tena tean tarn
tethra tether tethera
pethra mether fethera
pimps pip pubs
sarfra lezar aayther
larfra azar layather
ofra cartah quoather
dofra horna quaaather
dix dick dugs
ena dix yan-a-dick arnadugs
tena dix tean-a-dick tarnadugs
tethra dix tethera dick tetheradugs
pethra dix methera dick fetheradugs
bumpit bumfit buon
ena bumpit yan-a-bum arnabuon
tena bumpit tean-a-bum tarnabuon
tethra bumpit tethera bum tetherabuonf
pethra bumpit methera bum fetherabuon
sigit jiggit gun-a-gun

After Witty 1927: 47-8

Other examples may be viewed by clicking here

In an attempt to to establish the provenance of the scores Witty compares them with other counting systems, examples of which are listed below.

Germanic Languages
Celtic Languages
Anglo-Saxon Old Saxon Old Frisian Old Norse Gothic Erse Scottish Gaelic Welsh
an en en einn ains aon aon un
twegen / twa tuena / tua tweer / twa tveir twai da do dau / dwy
threo thria thre thrir threis tri tri tri / tair
feower fiwar fiwer fiorir fidwor cethir ceither pedwar / pedair
fif fif fif fimm fimf cuig coig pump / pimp
six sehs sex sex saihs se se cwech
seofon sibun sigun siau sibun seacht seachd saith
eahta ahto achta atta ahtau ocht ochd wyth
nigon nigun nigun niu niun naoi naoi naw / nau
tien tehan tian tiu tiu deich deich deg/dic
After Witty 1927: 49

He concludes that the scores are clearly not Germanic in character and that their similarity to the Celtic forms suggests they must be lexical survivals. However, M V Barry re-evaluating the situation forty years later challenges this view.

Barry's (1967) approach to the issue was to review the available material written about the scores. Having assembled a collection of 104 publications, eighty of which were written either as articles or in book form during the 19th century, he provides a summary of the facts known about these counting systems and examines the theories relating to their origin . The following is an outline of Barry's observations:


Use Number
To amuse children in the nursery
Children's counting rhymes in games
Counting sheep (though one of these is questionable)
Counting cattle
Counting stitches in knitting
Reckoning money

After Barry 1967: 22

Because it is seemingly impossible to trace anyone who uses or has used the scores for counting,. Barry is inclined to believe that they have been out of use from at least the mid-1870s.


To date, nothing has been conclusively established. Barry takes the view that the scores are linguistically linked to the Brythonic branch of the Celtic languages. Moreover, he considers that they contain Welsh elements (as opposed to Breton or Cornish). He bases this notion on the observation that all Celtic languages, with the exception of Welsh, follow the pattern: 1 to 10; then 1+10, 2+10, 3+10, etc. up to and including 19 (9+10). Welsh on the other hand takes the form of: 1 to 10; 1+10, 2+10 etc. up to 14; and then 5+10, 1+5+10, 2+5+10 etc. Compare the following:

Number Modern Welsh Millom, Furness score
1 un aina
2 dau peina
3 tri para
4 pedwar pedera
5 pump pump
6 chwech ithy
7 saith mithy
8 with owera
9 naw lavera
10 deg dig
11 (1+10) un-ar-ddeg aina-dig
12 (2+10) deudegg peina-dig
13 (3+10) tri-ar-ddeg para-dig
14 (4+10) pedwar-ar-ddeg pedera-dig
15 (5+10) pymtheg bumfit
16 (1+5+10) un-ar-bymtheg aina-bumfit
17 (2+5+10) dau-ar-bymtheg peina-bumfit
18 (3+5+10) deunaw para-bumfit
19 (4+5+10) pedwar-ar-bymtheg pedera-bumfit
20 ugain giggy

After Barry 1967: 27

If it is accepted that the scores are of Welsh origin the question remains as to why they are found in the areas mentioned above. Barry consider three theories: the Survival Theory, the Importation Theory Type 1 and the Importation Theory Type 2.


The Survival Theory
This is the earliest theory and suggests that the scores are a linguistic survival from the time when the Britons lived in the region. Barry notes that British place-names do exist in the areas where the scores are found which suggests the presence of isolated Celtic communities among the Anglo-Saxon and Scandinavian settlements. There is also some evidence of Celtic survival in the Lower Dales, the North Yorkshire Moors and on the Yorkshire Plain (Jones, 1965: 1966 cited by Barry).

However, there are a number of problems with this theory. inter alia:

  • There is a distinct lack of evidence for the existence of the scores prior to 1745.
  • Given that the scores are perceived as having a common origin,the differences between the various forms are difficult to explain.
  • Considering that little of the Brythonic language survives in place-names it seems extraordinary that it should survive in an entire sequence of numerals.

From the linguistic perspective, although nothing is known about Welsh numerals prior to the 13th century, the Welsh for 18 (deunaw) is well corroborated thereafter, whilst the employment of the form tri-ar-byntheg has been rare. Some manifestation of deunaw in the scores might be supposed if imported during the medieval or post-medieval periods. However, tri-ar-byntheg appears to be the basis of the customary form.


The Importation Theory, Type 1
This theory, for which there is no longer a great deal of support, proposes that Strathclyde Britons carried the numerals into the north-west of England. Whilst it is true that the former area was occupied the British and that, from the 13th century, droving roads passed through Carlyle and over to Penrith and Kirby Stephen, this theory encounters the same objections as those raised in relation to the Survival Theory. Indeed, Barry considers that the objections are even greater on the grounds that:


The Importation Theory, Type 2
This posits that the scores were imported into the north-west of England from Wales. Some landowners possessed estates in both Yorkshire and North Wales and droving from the former did exist from from the 16th century onwards. Accordingly, Barry proposes that labour could have been transferred between the two in both directions. As such, in-coming Welshmen and Yorkshiremen returning home could have brought the Welsh method of counting with them. Moreover, it has been suggested that Welshmen may have been attracted into the West Riding woollen industry following the downturn in the fortunes of their own at home. However, Barry was unable to uncover any evidence in support of such a notion and, in any event, such a situation would not explain the concentration of the scores around the North Riding and Furness areas.

There is a further proposal that contact may have arisen out of the employment of Welsh labour in the Yorkshire lead mines. Such a notion is supported by the fact that, of all the examples of the scores, those of Pately Bridge and Greenhowe Hill bear the greatest resemblance to Welsh. Notwithstanding, Barry is of the opinion that further investigation is required before any firm conclusions can be drawn regarding the Welsh origins of the scores.


Thus, as things stand, the scores may be a corrupt form of Welsh rather than a survival of a Brythonic counting system. However, it may be argued that, as Modern Welsh is directly descended from the language of the Britons then, even as a corrupt form of that language, the scores are part of our ancient Brythonic heritage.



Barber, C. (1993) The English Language - A Historical Introduction, Cambridge: Cambridge University Press.

Barry, M. V. (1967) Yorkshire Sheep-scoring Numerals. In Transactions of the Yorkshire Dialect Society 12, 67, 21.

Baugh, A C and Cable, T (1993) A History of the English Language, Fourth Edition, London: Routledge.

Burchfield, R.W. (Ed), (1989) The Oxford English Dictionary, Second Edition, Vols I and II, Oxford: Clarendon Press.

Crystal, D (1988) The Cambridge Encyclopedia of Language, London: Guild Publishing

Crystal, D (1995) The Cambridge Encyclopedia of the English Language, Cambridge: Cambridge University Press.

Hughes, G (2000) A History of English Words, Oxford: Blackwell

Jones, G. R .J. (1965) Early Territorial Organisation in Northern England and it s Bearing on the Scandinavian Settlement. In The Fourth Viking Congress, 1965.

Jones, G. R .J. (1966) The Cultural Landscape of Yorkshire - The Origins of our Villages. In Yorkshire Philosophical Society Transactions, 1966.

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

Serjeantson, M. S. (1935) A History of Foreign Words in English, London: Routledge and Kegan Paul.

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

Stalmaszczyk, P. (1997) Celtic Elements in English Vocabulary - A Critical Reassessment, Studia Anglica Posnaniensia 1997, XXXII: 77- 87

Wakelin, M. F. (1977) English Dialects: An Introduction, Revised Edition, London: The Athlone Press, University of London.

Witty, J R (1927) Sheep and sheep-scoring. In Transactions of the Yorkshire Dialect Society 4, 28, 41

Wright, J. (1898 -1905) The English Dialect Dictionary, 6 vols, Oxford: Oxford University Press