Project information

The aim of this project is to make a collection of exemplary regional Irish language dialects from the RTÉ Archives available to the public. It includes 200 recordings from twelve counties and they cover Gaeltacht areas as they are understood today as well as areas where the language has long since disappeared. The focus is on the period from the 1940s to the 1990s and includes some of the earliest Irish recordings from the RTÉ Archives.

62 hours of sound recordings were selected from the archive of which about 20% has been transcribed, standardised and rendered searchable on the site. Outstanding speakers were selected who reflect the diversity of Irish dialects found in the RTÉ Archives. Users of the site will recognise famous Gaeltacht storytellers such as Peig Sayers, Colm Ó Caoidheáin and Micí Sheáin Néil Ó Baoill, and they can listen to native speakers from former Gaeltacht regions such as Mícheál Ó Maoldhomhnaigh from Tipperary, Maggie McGiff from east Mayo or Jane Nic Ruairí from Tyrone.

The recordings on this site come from Irish language programmes that were broadcast on RTÉ, both Raidió Éireann and Raidió na Gaeltachta. Famous broadcasters of the time, such as Proinsias Ó Conluain, Séamus Ennis and Aindrias Ó Gallchóir can be heard doing fieldwork around the country, as well as interviewers from the early years of Raidió na Gaeltachta, such as Mícheál Ó Sé, Pádraig Ó Catháin and Timlin Ó Cearnaigh, as the station came of age.

The recordings are plotted on an interactive map which acts as a browsing aid for the user. The recordings have been dialectically transcribed and these are available below the audio player and are synchronised word for word with the audio track in a way that will help the user to follow the conversation. The transcriptions have also been standardised and this version is given as a guide wherever the dialectal transcription deviates from standard spelling.

One of the most notable features of this resource is the way in which dialectal versions of the same word are grouped together under the standard version. For example, when searching for ‘cnoc’, both ‘cnoc’ and ‘croc’ will be among the results. This feature allows the user to search for any Irish word, view the distribution of the dialectal versions of that word on the map, and then listen to the best speakers pronouncing those versions.

Project stages

1. Selecting the recordings

These recordings – digitised in recent years as part of the RTÉ Archives digitisation programme – were selected based on geographical distribution, speaker fluency, gender and sound quality.

The project aimed to achieve as wide a geographical distribution as possible, so as to include examples of dialects that were still spoken, if in terminal decline, in the mid-20th century, alongside recordings from areas where Irish is still spoken today. Often only one or two recordings were available from areas where Irish had already died out as a community language by the middle of the last century, places such as Omeath, Co. Louth (Annie Uí Annluain) and Newcastle, Co. Tipperary (Mícheál Ó Maoldhomhnaigh), which demonstrates the immense value of these items as both linguistic evidence and social history.

There is much more material to be found in the RTÉ Archives from areas where the Irish language was still alive and thriving during the period in question. In these cases a large number of recordings were assessed and those that best satisfied the aims of the project were selected. Geographical distribution was also taken into account here, but at a more granular level, with the aim of representing as many sub-dialects as possible within recognised Gaeltacht counties such as Donegal, Galway and Kerry. Particular attention was paid to those who are widely recognised as excellent speakers and valuable advice was received from broadcasters and dialectologists in this regard. A particular effort was also made to promote women’s voices in these areas, to compensate as much as possible for the over-representation of men in the archive as a whole.

In terms of sound quality, some recordings were excluded due to unsatisfactory sound quality.

2. Metadata

This is a collection of archival material that already had comprehensive information attached to it thanks to the RTÉ archives team. This pre-existing metadata was exported to the project database and was then enhanced in various ways, particularly in relation to recording additional information about the speakers. It was possible to draw on other archival sources such as the National Folklore Collection to locate some of this information, but new research was required to fill in some of the gaps in information regarding the age and location of ​​certain speakers.

3. Transcription

The Gaois research team transcribed each recording with the aim of providing a text version that would be faithful to the speaker’s dialect and easy to read for the average user. To this end, it was decided not to utilise a narrow transcription system (e.g. the International Phonetic Alphabet), because although this would be more accurate, few people other than trained linguists would understand it. Instead, a broad transcription system was used that is based on the system followed by the folklore collectors of the Irish Folklore Commission, a system that adheres to the orthography of the Irish language to represent pronunciation. The disadvantage of this approach is that there is no standard system for representing certain sounds. For that reason, it should be acknowledged that this transcription effort is a trial process and that it will be gradually adjusted based on both internal review and feedback from users of the site (to gaois@dcu.ie).

4. Standardization

Custom software was used to standardize the transcribed texts. As part of this process, different dialectal versions of the same word were grouped under the standard version, which significantly enhances the site's search functionality. In addition, the standard version is given in grey font below the dialectal version wherever the transcription deviates from the standard spelling, which should help the user understand variants they do not recognise.

5. Forced alignment

The next step in the process is to link the transcripts to the audio recordings so that each word in the transcript is synchronised with the appropriate section of the audio recording. This process makes it possible to jump from searching for a word on the home page to where that word is heard in a particular recording. This also enhances the experience of the person listening to the recording as the conversation can be followed word for word thanks to the highlighting of the current word.

6. Publication

Coming soon…

Offensive material

The selected extracts from the RTÉ Archives on this website reflect the evolution of broadcasting in relation to the Irish language over several decades, covering a range of genres, subjects and voices.

The recordings were selected with the aim of representing the widest possible spectrum of Irish dialects as found in the digitised materials. These include voices from recognised Gaeltacht and Breac-Ghaeltacht areas as well as areas where Irish is no longer spoken but where archive material exists.

As it is derived from recordings spanning a period of RTÉ broadcast history from the 1940s to the late 1990s, users should note that this content reflects the norms and values from the periods and the context in which the recordings were made. The content may contain isolated instances of language and negative stereotypes deemed offensive by contemporary standards. Expressions may no longer be considered appropriate, may use inaccurate terms or may be considered offensive language.

Should you require further information or clarification regarding material hosted on this site, or wish to express a view on the inclusion of a particular item, we welcome your comments. The editorial team can be contacted at: gaois@dcu.ie.