Study title

Wales

Individual

Organization

Family

Family: Household family

A sampling frame was created to underpin the data collection for the project, to ensure that we captured a range of different speakers across different discourse contexts and geographical locations. The sampling frame was designed to reflect current demographics of Welsh speakers to ensure that it reflects the contemporary sociolinguistic situation of the language as accurately as possible.Spoken data was sourced via two main approaches: (i) recruitment of participants to be recorded and (ii) recruitment of participants to contribute spoken data via a novel CorCenCC crowdsourcing app. The scope of (i) included not only research assistants going into the field to record speakers but also participants recording themselves in various interactions. This was facilitated through a network of local 'champions' (active language animateurs in targeted areas) or the Mentrau Iaith (each local authority in Wales has an associated Menter Iaith, i.e. community-based organisation dedicated to raising the profile of the Welsh language local language initiatives). Recruitment for (ii) was achieved by publicising the app (for example through social media, television appearances and publicity materials) to endeavour to reach a different cohort of participants who would be recording individually and in more private domains. Large Welsh language events such as the National Eisteddfod and Tafwyl provided opportunities for the team to reach a large cross-section of participants as well as raise general awareness of the project.The crowdsourcing app was made available on IoS, Android and via a web-interface, and campaigns in the media e.g. appearance on television programmes such as S4C's Prynhawn Da, on both Welsh and English medium radio and through local engagement events. Promotional material included pens, coasters, leaflets and postcard size information sheets. An 'unofficial' mascot - based on a cat called Cor-pws - was designed to facilitate the participation of those under 18 and proved popular with contributors of all ages. Facebook and Twitter accounts for CorCenCC were set up in the first months of the project to further enhance the recruitment and participation of contributors.Novel transcription conventions were devised for processing CorCenCC's spoken data (which was captured via the CorCenCC crowdsourcing app or manually, using audio recording devices). These conventions enabled us to fully reflect the whole spectrum of dialect/register variation captured in our speech data (making them more useful to academic researchers) as well as more accurately representing the speech of participants itself. In terms of written data, the good relationship forged at the beginning of the project with Welsh language publishers such as Gwasg y Lolfa led to the incorporation into the corpus of many up to date novels and books. A unique source of written data in the Welsh language is the locally based Papurau Bro (i.e. local community Welsh-language newspapers). Fairly rapid data capture, for example, sampling from the Welsh language academic journal Gwerddon through the Coleg Cymraeg Cenedlaethol and adult L2 pedagogical resources / examination papers through the Welsh Joint Education Committee resulted from our engagement with other project stakeholders in the planning process for the project.Regarding e-language data, website owners and blog authors cooperated generously and targets were exceeded. Contributors of SMS messages and emails were recruited in the same way as for the spoken data. All relevant participant information and descriptive metadata was recorded at the time of data collection. Permissions to share the data in an online public resource were essential to the development of CorCenCC. These permissions were obtained from the relevant legal entities (e.g. the copyright owner; the speaker themselves) before the data was collected and locally stored. The raw data together with the corresponding permissions and metadata were deposited into a local file storage system.