At every decade since 1891 the inhabitants of Wales have been asked if they can speak Welsh. Ever since, the question has been used to judge the health of the Welsh Language.
During the 20th century, the census showed a steep drop from just under one million Welsh speakers to around 500,000 in the 90s. But the number now seems to have stabilised.
Of course, despite ‘only’ halving, the number has continued to fall as a percentage as the Welsh population has increased.
However, one gets the feeling that the Welsh Government views this stabilisation as something of a victory in itself. At least the language isn’t losing ground.
The trouble, however, is that this census question has become increasingly useless and misleading as a way of measuring the health of the Welsh language, and is only going to become more so.
The problem is that the Census only asks if you ‘can’ speak Welsh, and nothing more. No question is asked about whether you actually use Welsh, whether it is your mother tongue or second language, or even how fluent you are.
One hundred, or even fifty years ago, this wasn’t a problem. Then, the vast majority of Welsh-speakers spoke it precisely because it was their mother tongue – the number of second language speakers was much lower back then.
Therefore the Census figures were very much an accurate indication of the numbers and percentages of people speaking it as their day to day language.
Not anymore. Since then we’ve seen a big increase in the number of second language speakers due to the rise in Welsh language education, and the collapse of Welsh as a mother tongue in areas where Welsh used to be the community language.
As a result, the overall figure in the Census is now wildly optimistic, and gives a very misleading impression of the health of the Welsh language.
Figures from the School Census of 2013, below, allow us to compare the percentage of people able to speak Welsh with the percentages of primary school pupils speaking Welsh at home.
Note that the 2011 Census figures refer to actual settlements, and not necessarily the electoral wards, which may have the same name but cover a smaller/larger area.
|Name of Town/Village||% of People ‘able’ to speak Welsh in 2011 Census||% of PS children speaking Welsh at home in 2013 School Census||% points difference, +/-|
What these five examples from Gwynedd show is that the 2011 Cenus is very slow to reflect the linguistic change that happens when a Welsh speaking community becomes anglicised.
Only in those communities where the Welsh language was strongest, like Llanrug and Llanuwchllyn, did the 2011 census give us an accurate indication of the situation on the ground.
In areas where anglicisation is occurring, or has occurred, the 2011 Census figure does a very good job at hiding it.
The census makes us most over-optimistic where we can least afford to be – and that’s precisely due to the way the question is phrased.
What is the Alternative?
The Census Question needs to be changed, and it’s obvious what the alternative is, for that alternative is the one used in most other countries in the world.
In Estonia, in Ethiopia, in Peru, in the Austro-Hungarian Empire, the main language question asked is/was ‘What is your mother tongue?’/ ‘What was the main language spoken to you at home during your childhood?’
In fact, the United Kingdom already does it with regards to immigrant languages; at the 2011 Census, every resident of Wales was asked if their home language was anything other than English or Welsh. That question needs to be modified so that Welsh and English are included.
There then needs to be a second question, asking everyone whose mother tongue is not Welsh, to state their level of ability in that language.
That way, we will have a much more accurate picture of who speaks Welsh and to what extent, and our efforts can be more effectively targeted in ensuring that the Welsh language is allowed to thrive in these communities.