Welsh Language
Next Selection Previous Selection
The Census Act, 1920, specifically mentioned language as one of the matters in respect of which particulars may be required to be stated in Census Returns, and provision was made in the Order in Council of 1931 for information on this subject to be stated in returns made in Wales and Monmouthshire. The Census of 1891 was the first at which an inquiry with regard to language was made, and the inquiry has been repeated at each subsequent Census. At each Census prior to that of 1931 the Census schedule provided for information to be stated thereon as to whether the person was able to speak English only, Welsh only, or both English and Welsh, but the Order in Council for the Census of 1931 omitted any reference to those able to speak English only, and consequently no inquiry on the point was included in the schedule for the 1931 Census. The omission of this particular category was deliberate, it having been included at previous Censuses not
with any intention of obtaining and publishing statistics respecting persons able to speak English and no other language, as might be implied from the form of the question, but rather with a view to making the questions as to ability to speak Welsh or not exhaustive and obtaining some
reply in the case of every individual. A person able to speak English but no Welsh who was, however, able to speak some foreign language, would have been justified in making no reply to such an inquiry since it could not truthfully be stated that English only could be spoken. The number of cases, in which no statement was made in reply to the language question was over 98,000 at the Census of 1921, and it is possible that a considerable proportion of these cases related, not, as might normally be expected, to persons who accidentally omitted to reply to the question, but rather to those who, though unable to speak Welsh, were able to speak some foreign language as well as English, or some foreign language only, and could not therefore enter any of the three terms of reply, viz., "English", "Welsh", "Both" in the Census schedule. The new form of question adopted at the 1931 Census sought to elicit merely the information that was required, viz., whether the person was able to speak Welsh only or Welsh and English, a dash (-) being inserted in respect of all children under three years of age and all other persons unable to speak Welsh. Thus, no attempt was made to make the questions exhaustive, the insertion of a dash being accepted as a proper response to the question and as signifying that whatever other language a person might be able to speak he or she could not speak Welsh. Some consideration was given to the question of furnishing guidance with regard to the degree of acquaintance with the language which would justify a person's claim to be able to speak it, but it was ultimately decided to leave the interpretation of the words "able to speak" to the discretion of the individuals responsible for making the returns and the statistics will accordingly be comparable, in this respect, with those of previous Censuses. The returns themselves will, as in the past, lack uniformity, some persons regarding the subject of inquiry to be the extent of the use of the language for the ordinary purposes of life whilst others would regard it as intended to elicit information as to proficiency in the use of the language from a literary or educational standpoint. It should be stated that the inquiry as to language was only made of persons who were enumerated on Census night in Wales and Monmouthshire and included, therefore, visitors and persons of other than Welsh domicile, whilst residents of Wales and Monmouthshire enumerated in England or elsewhere were excluded. As previously stated no return was required to be made with regard to children under three years of age, this practice having been followed at each Census since that of 1891, when only, from this inquiry, children under two years of age were excluded. Census schedules printed in Welsh were provided, as at previous Censuses, for the use of those householders who were unable to speak English and care was taken to appoint enumerators able to write and speak Welsh for duty in those parts of Wales in which such persons were likely to be enumerated. It was not contemplated, at this or any previous Census, that schedules printed in Welsh would be demanded for use by persons able to speak both English and Welsh, yet such was the case at the 1931 Census. The use of such a schedule serves no purpose in connection with the inquiry as to ability to speak Welsh since the statistics on this subject are derived solely from replies to the question which appears on all the schedules issued in Wales and Monmouthshire whether printed in the English or Welsh language. The fact that a particular householder fills up a schedule printed in the Welsh language is not recorded and is not the subject of any Census statistics, and compliance with the suggestion that the distribution of schedules printed in Welsh should be extended to individuals who, though able to speak Welsh, are equally able to speak English, would have involved an unnecessary burden upon the Census administration without any compensating advantage. In cases, therefore, where a concession was made and schedules printed in Welsh were supplied to individuals speaking both English and Welsh, it was made a condition that the return itself should be written in English in order that both the enumerator whose duty it was to examine and correct the schedule, and the householder whose duty it was to supply particulars capable of being understood, might be saved the unnecessary trouble that would have been involved if the enumerator had been unable to speak Welsh, Welsh-speaking enumerators having been appointed with certainty only in those districts where Welsh monoglots were expected. Statistics respecting the Welsh language are given in Tables 17 and 18, of Part I, of the series of County Volumes for Wales and Monmouthshire, and in Tables 39—42 of the General Tables Volume. Table 17 of the County series gives, by sex and groups of age, the total population and the numbers of persons able to speak Welsh only or both English and Welsh in the respective administrative counties, in each of the areas with populations exceeding 20,000 and in the combined areas making up the remainder of the county; Table 18 giving similar statistics, without distinction of age, in respect of administrative counties, municipal boroughs, urban and rural districts and civil parishes. The age group statistics above-referred to are summarised for the whole of Wales and Monmouthshire in Table 39 of the General Tables Volume whilst Table 40 of the same volume gives for Wales and Monmouthshire as a whole the proportions per 1,000 of persons in grouped ages returned as able to speak Welsh only or both English and Welsh at the Censuses of 1901, 1911, 1921 and 1931 respectively. Table 41 of the General Tables Volume gives for each county, county borough and urban area with more than 50,000 population, the total population, aged three years and over, and the numbers of persons, distinguishing males and females, in that age category able to speak Welsh only or both English and Welsh, whilst Table 42 gives for the same areas the proportion per 1,000 population aged three years and over of persons returned as able to speak Welsh only or both English and Welsh at the respective Censuses of 1901, 1911, 1921 and 1931. The returns show that of a total population of 2,472,378 persons aged three years and over, viz., 1,232,580 males and 1,239,798 females, 97,932 persons, viz., 48,629 males and 49,303 females claimed to be able to speak Welsh only and 811,329 persons, viz., 407,428 males and 403,901 females both English and Welsh. At the Census of 1921 the total population aged three years and over was 2,486,740, viz., 1,243,768 males and 1,242,972 females, of whom 155,989 persons, viz., 76,591 males and 79,398 females claimed to be able to speak Welsh only and 766,103, viz., 381,966 males and 384,137 females, both English and Welsh. From these figures it will be found that in a general population which has decreased by 63,142 persons or 2.4 per cent, since 1921 the number of persons returned as speaking Welsh only has decreased from 155,989 to 97,932, i.e., by 58,057 or 37 per cent., whilst the number returned as speaking both English and Welsh has increased from 766,103 to 811,329, i.e., by 45,226 or 6 per cent. A percentage decrease of 18 in the number of Welsh monoglots between the Censuses of 1911 and 1921 has thus become a percentage decrease of 37 between 1921 and 1931, whilst a decrease of 27 per cent, in the number of bi-linguists between 1911 and 1921 has become an increase of 6 per cent, between 1921 and 1931. The difference, viz., 1,563,117, between the figure of 2,472,378, which represents the number of persons three years and over in Wrales and Monmouthshire, and the total number of persons, viz., 909,261, able to speak either W7elshonly, viz., 97,932, or both English and Welsh, viz., 811,329, represents those persons in respect of whom no statement as to language spoken was entered on increase of 6 per cent, between the Census schedule. This figure of 1,563,117 therefore relates to persons (1) able to speak English only, (2) able to speak only some language other than English or Welsh, or (3) who failed to indicate ability to speak either WTelsh only or both English and Welsh. It corresponds closely with the 1921 figure of 1,564,648, which represents persons returned at that Census as able to speak English only, viz., 1,466,211, and those who failed to make any statement as to ability to speak English only, Welsh only or both English and Welsh, viz., 98,437. This latter number was made up of 79,816 cases of persons over 15 years of age who either justifiably failed to answer the inquiry or inadvertently omitted to do so (the respective proportions of these being unknown), together with 18,621 children aged 3—15 years in respect of whom no statement was made but in whose case the head of the household was returned as speaking English only, Welsh only, or both English and Welsh. No provision was made at the 1931 Census for identifying the children aged 3—15 who made no statement as to language, and it is not possible to determine the number who inadvertently omitted to reply to it, action taken being confined to eliminating those cases where in view of the terms of the question, no reply was possible. The proportion per 1,000 of the Welsh speaking population aged three years and over tabulated at each of the last four Censuses is shown in the following table :— From these figures it will be seen that there has been a progressive decline in the numbers of the Welsh speaking population since 1901, the percentage of those aged three years and over returned as able to speak Welsh only or both English and Welsh being respectively 49.9, 43.5, 37.1 and 36.8 at the Censuses of 1901, 1911, 1921 and 1931, the decline between the last two Censuses being appreciably smaller than those recorded in respect of previous intercensal periods. The decline has been most pronounced and most persistent in respect of those returned as able to speak Welsh only, Welsh monoglots comprising only about 4 per cent, of the total population as compared with about 15 per cent, in 1901. Bi-linguists at the Censuses of 1901 and 1911 comprised about 35 per cent, of the total population. Their numbers fell to about 31 per cent, at the Census of 1921 but at the 1931 Census recovered to about 33 per cent. The following table shows the age distribution of the Welsh speaking population giving for seven groups of ages the proportion which the numbers in the respective groups bear to the total number living in those groups. From the above figures it will be seen that, as at previous Censuses, the age incidence of persons speaking Welsh only is relatively high at the earliest ages, viz., 3 to 9, decreasing rapidly to a minimum at ages 15-24, remaining almost constant at ages 25-44, and thereafter rising until at ages 65 and upwards it is appreciably higher than at any other period of life. The decrease in the numbers speaking Welsh only as compared with 1921 is spread over the whole of the age groups but the proportional decrease is greatest at ages 25-44 amounting to as much as 50 per cent., whilst at age groups 15-24 and 45-64 the proportional decreases are 44 and 45 per cent. respectively. The lowest decrease is at ages 3-4, viz., 14 per cent. The frequency of bi-linguists will be seen to be highest at ages 45-64, the proportion of persons speaking both Welsh and English rising from 125 per 1,000 at ages 3-4 to 396 per 1,000 at ages 45-64. The increase between 1921 and 1931 in the numbers of bi-linguists is not spread over the whole range of age groups as was the decrease in the numbers speaking Welsh only, age groups 3-4 and 5-9 showing a proportional decrease of 19 and 5 per cent, respectively, whilst increases range progressively from 2 to 11 per cent, in the remaining age groups, other than that of 10-14 which remains constant. The distribution of the Welsh speaking population has hitherto been discussed for Wales and Monmouthshire as a whole but in the following table the distribution is extended to the individual counties and to the large towns. The areas are arranged in the order of the frequency of the total Welsh speaking element in the population, and comparable figures for 1921 are included with those for 1931. The decrease of 58,057 in the number of persons able to speak Welsh only was shared in varying degrees by each of the thirteen counties, proportional decreases varying from 64 per cent.—the maximum—in the case of Glamorgan and 56 per cent, each in the case of Flint and Brecknock to 18 per cent.—the minimum—in the case of Caernarvon. The greatest numerical decrease, viz., 16,411 occurred in respect of Glamorgan followed by Carmarthen—11,335 - and Caernarvon—7,828-, these figures comparing with reductions between 1911 and 1921 of 9,396— the maximum—in Caernarvon, 6,179 in Glamorgan and fewer than 5,000 for any other county. The proportion of persons speaking Welsh only to the total Welsh speaking population was less than one ninth in 1931 as compared with about one sixth in 1921. With regard to persons able to speak both English and Welsh there was an increase in numbers between 1921 and 1931 of 45,226 as compared with a decrease of 20,971 between the Censuses of 1911 and 1921. The increase was spread over ten of the counties, three showing decreases of small magnitude. The largest increase, viz., 16,986 was shown by the county of Carmarthen which also had the largest increase, viz., 11,893 between 1911 and 1921. Denbigh showed an increase of 6,805, against a decrease of 5,050 in the previous decade, and Caernarvon an increase of 6,038, whilst three counties had increases varying between 3,000 and 4,000, two between 2,000 and 3,000, and two of less than 1,000. Glamorgan which showed between 1911 and 1921 a decrease of 18,616 compared with an increase of nearly 70,000 between 1901 and 1911, registered a small increase of 2,887 between 1921 and 1931 The geographical distribution of the Welsh speaking population will be seen to have been predominantly in the five counties, Anglesey, Carmarthen, Cardigan, Merioneth and Caernarvon, in that order of frequency in 1921 The same counties with a slightly varied order of frequency had, in 1931, a high proportion of their population, varying from 79 to 87 per cent , able to speak Welsh, either alone or in conjunction with English At the other end of the scale are the counties of Radnor and Monmouth with populations speaking almost exclusively English, the Welsh speaking element comprising only about 6 per cent, and 5 per cent, respectively of the population of those counties The counties of Denbigh and Montgomery again occupy an intermediate position, the Welsh speaking populations comprising between 40 and 50 per cent, of the total, whilst in the remaining counties of Brecknock, Flint, Pembroke and Glamorgan the language spoken is predominantly English, between 30 and 40 per cent, only speaking Welsh Table XIV of the Text portion of the respective County Parts (Volume I) gives proportion per 1,000 of the population aged three years and over in urban and rural districts returned as able to speak Welsh only and both English and Welsh, with comparative proportions for the two previous Censuses From the figures contained therein it will be seen that, in some of the rural areas particularly, the proportion of Welsh monoglots is high compared with the figure representing the county proportion in the table above The highest proportions of persons speaking Welsh only were found in: the proportions in respect of these areas in the same order having ranged from 573 to 512 per 1,000 at the 1921 Census The highest proportions of persons speaking Welsh, viz , monoglots and bi-linguists combined were found in: The highest figure, viz , 973 is in respect of an area situated in Pembrokeshire in which county only 306 per 1,000 of the population as a whole are able to speak Welsh. PART VII — WELSH LANGUAGE
1. Scope of Inquiry.
2. Age Distribution.
3. Local Distribution of the Welsh Speaking Population.
Penllyn Rural District (Merioneth)
493 per 1,000
Lleyn Rural District (Caernarvon)
486 per 1,000
Uwchaled Rural District (Denbigh)
474 per 1,000
Glaslyn Rural District (Caernarvon)
445 per 1,000
Gwyrfai Rural District (Caernarvon)
414 per 1,000
Dwyran Rural District (Anglesey)
401 per 1,000
Llanfyrnach Rural District(Pembroke)
973 per 1,000
Aberayron Rural District(Cardigan)
959 per 1,000
Newcastle in Emlyn Rural District(Carmarthen)
958 per 1,000
Uwchaled Rural District(Denbigh)
948 per 1,000
Tregaron Rural District(Cardigan)
947 per 1,000
Machynlleth Rural District(Montgomery)
944 per 1,000
Lleyn Rural District(Caernarvon)
943 per 1,000
Gwyrfai Rural District(Caernarvon)
941 per 1,000
Llandyssil Rural District(Cardigan)
937 per 1,000
Dwyran Rural District(Anglesey)
936 per 1,000
Twrcelyn Rural District(Anglesey)
936 per 1,000
Llanybyther Rural District(Carmarthen)
935 per 1,000
Lampeter Rural District(Cardigan)
934 per 1,000
Bethesda Urban District(Caernarvon)
933 per 1,000