Well, you’d think so – so I appreciate your surprise, coming from a place of greater knowledge than mine – which is why I double-checked before commenting
The full context is this:
I Iorwerth Peate ac eraill roedd dweud tri deg saith yn lle dau ar bymtheg ar hugain am y rhif 37 yn gwbl annerbyniol ac yn enghraifft o 'siarad Saesneg yn y Gymraeg'. Rhoddodd Iorwerth Peate y bai yn blwmp ac yn blaen ar ysgolion fel y 'prif bechaduriaid yn y proses o ddileu'r dull traddodiadol Cymraeg o gyfrif mewn ugeiniau'. Datblygodd y ffrae yn un gyhoeddus yn ystod y 1970au. Mewn llythyr yn y wasg, gofynnwyd i Iorwerth Peate beth oedd forty-eight thousand (48,000) yn Gymraeg. Daeth yr ateb wyth mil a deugain [my emphasis] yn syth ganddo. Os felly, mentrodd yr amheuwr, beth yw eight thousand and forty (8040)? Ni chafwyd ymateb a bu taw ar y ddadl.
Rough Englishing for the sake of Forum protocol:
“For Iorwerth Peate and others, saying tri deg saith instead of dau ar bymtheg ar hugain for the number 37 was completely unacceptable and an example of ‘speaking English in Welsh’. Iorwerth Peate put the blame fair and square on schools as ‘the chief offenders in the process of destroying the traditional Welsh way of counting in twenties’. The fight became public in the course of the 1970s. In a letter in the press, Iorwerth Peate was asked what forty-eight thousand was in Welsh. His answer came straight back - wyth mil a deugain. In that case, ventured the sceptic, what is eight thousand and forty? No answer was received, and the argument went quiet.”
ETA: I’ve just checked Gareth Roberts’ list of traditional vs decimal counting in his appendix, and it’s interesting: for 28 he gives wyth ar hugain, but for 48 it is wyth a deugain. Hmmm.
Except that Roberts’ appendix consistently uses a rather than ar from deugain onwards, as does Wicipedia (perhaps not such an august and dependable source) and my old (Literary) Modern Welsh Grammar (Stephen J Williams), which gives dau dudalen a deugain for ‘forty two pages’ and un mlynedd a deugain as examples. And you yourself give deg a deugain in my other grammar (but no other examples I can find)… so it looks like the system changes from ar to a from deugain onwards(?)
I am so glad I started this thread. It’s been fascinating (if a little confusing at times!)
From a position of total ignorance plus a dash of mathematician’s logic, I would have guessed that the distinction would be 48,000 = wyth ar ddeugain mil and 8040 = wyth mil ar ddeugain .
Evidently the correct answer is even more subtle (and pretty tricky to pick up aurally).
Interested to know the answer to Richard’s point too, as to when and where to use ‘a’ vs ‘ar’
OMG so I did! Well I suppose therefore that I must indeed have been following other authorities for this (NOT Wicipedia, mind!) - probably Williams who is jolly good for LitWelsh, and deg a deugain is certainly a LitWelsh form very rarely heard in speech, where it’s always hanner cant.
In that case Peate was NOT wrong (although it’s odd that he didn’t carry the discussion any further) - but then if that is so, can we plausibly accept a numeral system that includes direct ambiguity? That would be highly unusual in any language, for obvious reasons.
I suspect that there were relatively few circumstances under which this ambiguity would have actually arisen under practical circumstances at the time that the system was actually evolving – unless you were facing a force of, say, 48000 Saxons, and had stopped to count them…
I did, in the end, get as far as checking the old Middle Welsh Grammar, too – on the basis that William Morgan’s usage was already literary and archaic even in his day, so why not? – to find that D Simon Evans makes the distinction explicit: “from 20 to 40, 1, 2, &c., are placed with ar before […] hugein(t) […] and 40, 60, 80, &c. are connected by a(c) ‘and’, thus: un a deu ugein(t).”
@Catriona – It’s going to work like “threescore years and ten” in English – “threescore and ten years” just sounds… wrong.
What this DOES show (and despite my staunch defence of the lower and commonly-used traditional numbers, I do agree with this) is that the decimal system is obviously superior for very high numbers: 48,000pedwardeg wyth mil, vs 8,040wyth mil a phedwardeg.
And I therefore stick obviously with my remark that in that exciting grammar that they’re really not used much above 50.
But we must keep saying deuddeg o’r gloch and ugain munud and pymtheg milltir and the like - because that’s what (many of) the native speakers say!
Yes but (mathematician speaking here, not Welsh learner) that’s kind of the point. “Three score years and ten” = 320 + 10 = 70. Whereas “three and ten score” would sound more like 13 score or 260. While “ten score and three” would be 1020 +3 = 203
Now somebody (probably Catriona, I hope) is going to ask: where is the dividing line between <numeral + singular noun> and <numeral + o + plural noun>?
And I will give the same answer that I gave elsewhere: a line running east-west through Y Ffarmers pub in Llanfihangel-y-Creuddyn.
Well, I was struck by the direct ambiguity, but I freely admit that introducing it here (and ‘atting’ you) was a matter of… teasing? Baiting the bear in his den? – when what is actually needed is a recognition that both systems currently exist and should be understood – and, ideally, used appropriately according to context. One thing I do think is interesting in the Gareth Roberts book is the extent to which he recognises this code-switching with numbers – something I suspect is cross-linguistically perhaps a fairly unusual phenomenon…
Thank you for this fascinating post @Catriona - I had no idea there was a whole different counting system that has survived decimalisation. Cymraeg just gets more and more interesting! I look forward to trying to get my head around it all one of these days.
