When I was a student at Queen's University Belfast towards the end of the 1960s, I became aware from some of my friends who were studying Irish there — in particular from Dr Eugene McKendry — that the library of the Department of Celtic possessed a box of records, and that these records contained spoken Gaelic from many different parts of Ulster. In my experience up until then, spoken Ulster Gaelic had been synonymous with the language of Donegal. Everyone knew, of course, that Gaelic had once been universally spoken in Ulster, and that Ulster was located at the geographical and linguistic centroid of the Gaelic world. We were even familiar with the names of individuals from various parts of Ulster whose speech had been phonetically transcribed early in the twentieth century by linguists like Sommerfelt, Holmer, Ó Searcaigh and Ó Tuathail. But the idea that we might actually hear the voices of such people had seemed like an impossible dream. Clearly here was a treasure, and I wanted to know more about it.
To actually play the records was not something I even considered. I might have considered it, if there had been a suitable and well-maintained apparatus in the University. I could have brought in a record player of my own — the records were fairly standard at 12 inches in diameter, 78 rpm, single-sided — but I was not prepared to risk damaging them. It was rumoured that some of them were damaged or broken already, probably in the course of being played by those who had a better academic right to hear them than I had. But there was one thing I could do, and that was to look for documentation.
Here I was in luck, up to a point anyway. The recording scheme had been administered by the Royal Irish Academy (RIA) in Dublin, and was described in their contemporary proceedings and minutes, and these were available in the Queen's University Library. I soon put together a folder of photocopied documents. The recordings had been made by a joint team of experts — Celtic scholars from Irish universities acting for the RIA, and sound recordists from the Lautabteilung of the Preussische Staatsbibliothek in Berlin — in the course of three visits by the latter to Ireland. The Munster recordings were made in September 1928 (serial numbers LA1031–LA1092); the Connacht recordings in September 1930 (serial numbers LA1103–LA1177); and the Ulster recordings in September 1931 (serial numbers LA1201–LA1281). Professor Wilhelm Doegen, the head of the Lautabteilung, attended in person only in 1928, and was represented by an assistant, Herr Karl Tempel, on the other two occasions. It is therefore a slight misnomer to give the Doegen name to the whole project, but that is how it has come to be known.
As well as many administrative details, my sheaf of photocopies contained a lengthy report of the Munster work, and briefer reports of the Connacht and Ulster work, with in each case a list of the names and addresses of the speakers, and the titles of the items contributed by them on each record. In Ulster, 80 records were made, from 40 speakers (note that serial numbers LA1201 and LA1202 are parts of a single disc). A record could last up to about four and a half minutes, and might contain as many as five or six tracks (as well as a reference track containing a brief tone of known pitch). It further appeared that three physical copies of the records had been made. One copy had been kept in Berlin; one copy in the RIA in Dublin; and the third copy was distributed among the provincial universities, with the Munster records going to University College Cork, the Connacht records to University College Galway, and the Ulster records to Queen's University Belfast — the copy of which I had become aware.
The report of the Ulster project stated: "[written] transcripts of the matter recorded were made by secretaries before the speakers left the recording rooms". I did not get access to this material until later. When I did, I found that these written versions were not close copies of the sound recordings; the wording was often different, and occasionally an item which was incomplete in the sound recording was finished out in the written version. It is better not to call them transcripts but I will use the term "contemporary written version", or in Irish "ath-innse". There is a simple explanation for the differences. The sound recordings could not be played back immediately, but only much later, after the records had been manufactured from the matrices. So written versions could only be made on the spot by getting the speakers to repeat their contributions independently.
We can easily imagine the pressure under which these contemporary written versions must have been created. This will account for the lack of consistency in them, and for the occasional mistakes. It is best to regard them as field notes, which were — in most cases — never rigorously written up. Understood in this light, their lack of precision is not a serious drawback, but they provide valuable alternative verbal formulations of many of the contributions. Not least in importance, given that the transcribers had the opportunity to question the speakers, they give unique clues to the meaning of some obscure passages which might otherwise be forever beyond comprehension.
The RIA also made a limited early attempt to transcribe the Ulster sound recordings directly, soon after the records were received. The Minutes of Proceedings for Session 1933–34, dated 16 March, stated "complete transcripts, text, phonetic script, and translation of 6 of the Ulster records have been prepared." The six records were LA1205 (from Cavan), LA1211 (from Tyrone), and LA1228, LA1233, LA1247, LA1275 (all from Donegal). The transcribers appear to have been Éamonn Ó Tuathail and Séamus Ó Néill. I had no access to this material until later.
Séamus Ó Néill was one of the "secretaries" who had made the contemporary written versions of the Ulster material — he seems to have made most of them — though his name is nowhere mentioned in the RIA documentation. In his capacity as a journalist, he went on to publish two series of the pieces as examples of Ulster folklore, one series in the Irish Independent (4 Feb 1932 – 21 Apr 1932) and the other in An tUltach (1933–35), omitting to note either the RIA or the Doegen scheme as their source, though he had made some earlier references to the scheme. Ó Néill's versions are sometimes closer to the sound recordings and sometimes closer to the contemporary written versions. His Independent pieces have been republished in Nollaig Mac Congáil agus Ciarán Ó Duibhín, Glórtha ón tseanaimsir: Doegen agus Béaloideas Uladh, 2009.
Little further work was done on the Ulster Doegen recordings until 1962 when Colm Ó Baoill wrote an MA thesis entitled Phonetic Texts of East Ulster Irish at Queen's University under the supervision of Heinrich Wagner. For this thesis Colm transcribed the "East Ulster" recordings, comprising 30 records from 14 speakers, both in phonetic script and in normal orthography, and he added English translations (except for the several recitations of the numerals and of the days of the week, which are given in phonetics only). One record, LA1224, from Crossmaglen, is not covered; it is said to be "lost", but fortunately it seems to have turned up since.
Colm did not work from the records directly but from "a tape made by the staff of the Irish Folklore Commission from thirty record discs". Presumably this tape was made from the Belfast copy of the records, which seems to have been the only copy available to Colm; he mentions the Berlin copy but did not have access to it; he does not mention the Dublin copy. And it is unlikely that LA1224 would have been misplaced in two different copies. Even with the help of the tape, the conditions under which these transcripts were prepared must have been extremely difficult by modern standards. It is therefore no surprise that Colm has to acknowledge "many gaps" in both the Irish transcription and the English translation. Colm was helped somewhat by having access to the contemporary written versions, "separate and different" though they were.
Ó Baoill's transcriptions of the sound recordings were republished in Wagner and Ó Baoill, Linguistic Atlas and Survey of Irish Dialects, Vol IV, 1969, Appendix II. Omitted on this occasion were any songs or poems, some of the stories which had proved most difficult to interpret, and some of the Tyrone pieces which had meanwhile been published (though not from Doegen material) in Stockman and Wagner, "Contributions to a study of Tyrone Irish", Lochlann 3, 1965, pp 43–236.
Apart from transcribing the content of the recordings, the other main requirement was to gather information about the speakers. As already noted, their names and addresses had been published. Further, the lengthy report on the Munster recording session stated that "when the record had been taken, a detailed questionnaire (supplied by Dr Doegen), containing information as to his age, travels, parentage, etc., which will be very valuable for future reference, was filled up in duplicate from his statements by one of the experts. One copy was given to Dr Doegen, the other retained by the Sub-Committee [in the RIA]." The questionnaires were not mentioned at all in the much briefer Connacht and Ulster reports. I had as yet no leads to the existence or location of either copy of these questionnaires.
The state of affairs in Berlin was still unknown to me. But a colleague, Dr Rosemary Power, gave me a letter which she had received in December 1989 from Dr Martin Rockel of the Humboldt University, confirming the existence there of a collection of Irish language records, whose quality was described as variable, and which had not been academically worked on. A two-page article from Dr Rockel followed, in January 1990, from which it was clear that the collection contained the bulk of the Doegen records from all three provinces (though it was not possible to be sure of details). If I understand the paper correctly, it says that copies of the records were held by the Humboldt University before the war — the same copies, I think, that they still have; but the whereabouts of the "Exemplare" (originals?) was by this time unknown. The matrices were destroyed in the Lindström factory during the war. Dr Rockel was aware of the contemporary written versions, but did not know their whereabouts. He did not mention the speaker questionnaires. I exchanged letters with Dr Rockel in 1990/91, but I was not then in a position to progress matters further.
In August 1990 I wrote to Dr Colm Ó Baoill at the University of Aberdeen, and he immediately sent to Queen's University his personal copy of the contemporary written versions. This copy had a unique value due to the notes which Colm had made on it in the course of trying to bring it closer to the spoken texts. He had no knowledge of the speaker questionnaires, or of the careful transcripts.
By late 1993, it seemed that the time was due to digitise the Ulster Doegen records, both to safeguard their content and to make them easier of access. The Professor of Celtic at Queen's University, Gearóid Stockman, readily agreed, and we sought and received the support of the University's Institute of Irish Studies and of its director, Dr Brian M Walker. The project proposed not just to digitise the material but to disseminate it afterwards — cassettes were considered at the beginning, later changed to CDs; but as it turned out, the potential financial and legal problems raised by dissemination never actually came to a head, and have now receded with the rise of the internet as a publishing medium. It was also intended that I should provide a transcription of the texts, and should research the facts concerning the speakers, in a book to accompany the recordings, but this proved over-ambitious. I had obtained unpaid leave of absence from my post in the Department of Computer Science at Queen's University from February 1994 until February 1996, and this enabled me to accept a secondment to work on the Doegen project in the Institute of Irish Studies. Dr Walker was successful in obtaining financial support for the project from the Irish Soldiers and Sailors Fund of the Taoiseach's Office, to cover the cost of the sound transfer and to pay me a grant.
For guidance in sound transfer, we turned to the National Sound Archive (NSA), a part of the British Library. Dr Walker wrote to their conservation manager, Peter Copeland, who took a genuine interest in the project and expressed a willingness to part-fund the transfer in return for a copy of the material (we earmarked this recoupable part of the transfer costs for reuse to finance eventual publication). He recommended Ted Kendall, a freelance sound engineer living in Herefordshire, to carry out the work. The first contact with Ted Kendall was made in March 1994, costings were obtained, and it was decided that two digital sound versions would be produced: one, for archival purposes, would be as direct a transfer from the analogue discs as possible, and the material would remain in the order of the records; the second version would be aurally "enhanced" insofar as available techniques allowed, and would be rearranged into a geographically-logical sequence, devised for distribution over 3 C90 cassettes. The enhanced version would contain only one copy of each track, whereas the archive version would contain the Belfast and Dublin copies of each track separately.
In April 1994 Dr Walker contacted the secretary of the RIA, Paddy Buckley, to formally request permission to work on the Ulster Doegen material and to disseminate the results of such work. He also requested details of any holdings by the RIA of Ulster Doegen material: the actual recordings, the contemporary written versions, the speaker questionnaires, and any careful transcriptions. The Council of the Academy gave their permission in September 1994, and we learned from the RIA Librarian, Siobhán (O'Rafferty) Fitzpatrick, that there was a set of all 80 Ulster records in Dublin; that there were two folders of transcripts which turned out to contain another copy of the contemporary written versions, and also the careful transcipts of LA1205, LA1228, LA1233, LA1247 and LA1275; but there were no speaker questionnaires. The RIA strongroom also yielded a photograph of one of the recorded speakers, Barney McAuley of Glenariffe, Co Antrim, and the Librarian generously provided me with a copy. This photograph has become iconic of the recordings on the RIA website.
We had not been sure up to this point that a copy of the records existed in Dublin — it is nowhere stated in published documentation — but we had been assured by Professor Gearóid Mac Eoin that he knew of such a copy. In fact, the Dublin copy was in excellent condition, and the RIA agreed that it made sense to take both the Belfast and Dublin copies for digitisation. By comparison, it had emerged that two Ulster records of the Belfast copy were broken, while others had suffered "little wear but rather a lot of surface damage" according to a preliminary assessment by the NSA. After repairs it proved possible to play and transfer all the records of the Belfast copy also (though LA1201/2 was unavailable at this time and was deferred until later).
By 15/01/1995 I could tell Ted Kendall that the finance was in place, and that both copies of the records were cleared for copying, and so we arranged to carry out the work in mid-May, and that I would stay over and observe the process. When the time arrived, I collected the Belfast copy of the Ulster records from Queen's University in my car-boot, drove to Dublin and collected the Dublin copy from Academy House on 12/05/1995, and, acutely aware of the saying about putting all one's eggs in one basket, I caught the car-ferry from Dún Laoghaire to Holyhead, and drove to Gladestry in the Radnorshire–Herefordshire borders. All went well with the transfer, and I returned to Ireland with my fragile cargo on 20/05/1995, to await delivery by post of the digitised versions when the processing had been completed — one enhanced copy on which to base our publication, and three archive copies for the RIA, QUB and NSA.
Over the summer of 1995, the enhanced version arrived, one cassette's-worth at a time but actually on digital audio tape (DAT), for me to check as to completeless, sequencing, and general correctness. As I had difficulty obtaining the use of a DAT player, I had the DATs copied to cassette at Queen's University. I was at length able to report to Ted Kendall on 05/02/1996 that sequencing and completeless were perfect, but there appeared to be some unexpected variation in volume between tracks, and occasional transient noises, so that a new issue of the enhanced version would be needed. Meanwhile I was kept occupied with the twin tasks of transcribing the material and researching the speakers. My term of secondment to the Institute of Irish Studies was now rapidly coming to an end, and it was obvious that neither task would soon reach an acceptable level of completion to be placed before the public. In the course of the year I gave an Institute seminar entitled "Records of spoken Irish from the 1930s" (around 13 March 1995).
Henceforward, Ted Kendall would send everything on CD, and at the beginning of 1998 he sent me the latest enhanced version (on 6 CDs, I think, but destined to fit on 4), together with three copies of the archive version (each on 9 CDs). It was October 1998 before I managed to reply, and alas, with all three copies of the archive version, the same handful of tracks featured false starts, in which the first few seconds were repeated. The enhanced version was technically faultless, but by now our preferred distribution medium had changed from C90 cassette to CD, and, the length of a CD being 77–80 minutes rather than 90, the enhanced material would have to be divided over 4 CDs rather than 3 cassettes, and some rearrangement was necessary to keep each CD to some extent self-contained.
Before another delivery could be organised, events moved again. The RIA had decided to digitise their Munster and Connacht Doegen records also, and their librarian Siobhán Fitzpatrick asked me if I would bring these records to Ted Kendall also, which I was delighted to do. This project was instigated and financed by the RIA alone; there were no Belfast records this time, and no involvement of Queen's University. The records came from the RIA strongroom in Dublin, and for the Munster records, this Dublin copy was complete, except for the four records which were never made due to broken matrices (LA1048, LA1049, LA1055, LA1076). No copies of the Munster records were sought in Cork.
For the Connacht records, the Dublin copy appeared to be a mixture of two copies, distinguished by one copy having the record labels and sleeves stamped with the record number, while the other (very incomplete) copy had handwritten identification istead of stamping. For some records there were both stamped and unstamped copies, for some there was only one or the other, and for quite a few there was neither. Almost all the gaps were made good by records obtained on loan from Roinn na Sean-Ghaeilge in UCG — all of which were stamped. I do not know whether we received all the records available in Galway, or only those which the RIA requested, but in fact a few records were present in both a Dublin copy and a Galway copy. All the assembled copies of every record were digitised. Between the Dublin and Galway sets, only two records were still problematic. LA1108 was present in the Dublin copy but was unplayable; it proved possible however to repair it. Finally, no copy of LA1147 was available to us at this time.
The Connacht and Munster records were digitised during a visit by me to Gladestry which extended from 23/07/1999 to 29/07/1999, again using the car ferry from Dún Laoghaire to Holyhead. The objective was to make one enhanced and several archive copies, all for use by the RIA. No publication was planned at this stage. Some Dublin Munster tracks exhibited faulty sections which were digitised again under different groove wall parameters (LA1039d1, LA1069d1, LA1070d1, LA1071d1, LA1072d1, LA1080d3, LA1083d1). Advantage was taken of this 1999 visit to make the final minor adjustments to the Ulster digitisation: the addition to the archive version of the Belfast copy of LA1201/2, now rediscovered; removal of the false starts in the archive version; reordering of the enhanced version for 4 CDs.
On my return I had with me the archive versions of the Ulster, Connacht and Munster recordings. It is most likely that I had one copy of all three provinces for RIA, and extra Ulster copies for QUB and NSA, probably on high-quality gold CD blanks. I now renewed my attempts to contact NSA, to establish whether they were still interested in partially re-imbursing us in return for their Ulster archive copy, but a great deal of time had elapsed and no response was now forthcoming. In fact I still retain that original gold copy of the Ulster archive material. There would have been no reason for me to delay forwarding the other archive copies.
I cannot be sure when I received the enhanced versions of the three provinces (after the necessary re-ordering of the Connacht and Munster tracks), nor have I any record of when I despatched them to RIA (all three provinces) and QUB (Ulster only). I still retain a gold copy for all three provinces, and so probably made copies of these to normal CD blanks for distribution. I find myself informing Dr Walker on 26/09/1999, in relation to the Ulster material, that all had been received — both archive and enhanced — and, except for NSA, all had been forwarded. It later emerged however that the Ulster archive CDs were not received by RIA at this time, something which I am at a loss to account for.
New developments began when I noticed, in February 2000, that some new webpages had appeared on the internet concerning the Lautarchiv, mainly authored by Dr Cornelia Weber of the Humboldt University. I e-mailed Dr Weber, inquiring about the condition of the Ulster Doegen records and their documentation. Dr Weber passed my inquiry to Herr Jürgen Mahrenholz of the Lautarchiv, who replied on 07/03/2000, confirming (to my great relief) the existence of the speaker questionnaires, though their extent was still unclear. He also indicated that the records spanned the full range of serial numbers, though he remarked that many were damaged or missing. We quickly established that the only record of which no copy could be found in Ireland, LA1147, was present and in good condition in Berlin.
In March 2002 I informed the Secretary of the RIA of these contacts, and they were taken forward thereafter by the Librarian. I was very pleased to receive from Ms Fitzpatrick, by e-mail in October 2002, a copy in mp3 format of the missing track, LA1147; and in February 2003, a copy of the Ulster speaker questionnaires followed.
For some reason which I cannot now understand, it was only in May 2002 that I sent a copy of the 4 Ulster enhanced CDs to the RIA. At the same time it emerged that they had never received the 9 Ulster archive CDs, so a copy was made of these also and sent on soon after.
We had now reached the point where the recordings were available for the public to listen to, though only to personal callers at the Library in Academy House. All known documentation could also be examined in the Library. There was also a copy of the Ulster material readily and securely accessible in the Department of Celtic at Queen's University. And from my personal point of view, I had my own copies of the Ulster recordings and documentation to work with.
My work on the transcription of the Ulster material, and my research on the speakers, were proceeding whenever time allowed, but the intention of producing these in book form to accompany the Ulster recordings on sale had receded as a possibility. It mattered little therefore that we had not secured the partial re-imbursement of the sound transfer costs which was intended to contribute to the costs of publication. It now seemed more realistic to publish the matter on the internet — something which we could do while it was still incomplete. I put this idea to the RIA secretary in March 2002.
With technological developments, it was now an obvious step to make the sound material itself, as well as the documentation, available on the internet. The RIA Library made a proposal to the Digital Humanities Observatory project in 2007, funding was secured, Dr Eoghan Ó Raghallaigh was taken on as a post-doctoral researcher with advice from Professor Ruairí Ó hUiginn agus Dr Seán Ua Súilleabháin, and the RIA's Doegen web site, under the management of Siobhán Fitzpatrick, has now become a reality.
At the same time, the present web pages bring to a degree of fruition my own studies concentrated on the Ulster Doegen material, with the object of securing the future of the sound recordings and making them available with supporting information. The transcriptions, notes and speaker details offered here are, of course, not final, and may never be. My thanks are due to everyone who is named in the above saga, and probably to many who have not been named; if anyone has been forgotten, they have only to remind me and they will be speedily credited. I hope that these web pages will be acceptable to all who have supported this work as justification for that support.