Hildegard L. C. Tristram

• Periphrastic DO constructions are very common both in English and in the Neo-Brittonic languages and are used for various functional purposes. These form part of a larger linguistic area in western and northern Europe. The literature does not mention comparable constructions for Irish and Scottish Gaelic. Irish informants, however, confirm orally that they are in common use among present day Gaeltacht speakers. They appear also to have been common in late spoken Manx. This study is based on the "Caint Chonamara" electronic corpus, the field work for which was first untertaken by Hans Hartmann (Hamburg) and Tomás de Bhaldraithe (Dublin) in the early 1960s und brought to a close by Arndt Wigger (Wuppertal) in the 1990s. The file for the Ros Muc dialogues yielded a very low return of potential DO constructions, i.e. 14 tokens of DÉAN + VN out of 494 DÉAN tokens altogether in the file. This shows that the DÉAN + VN construction was grammatically correct and acceptable to the native speakers, but was not grammaticalised and had a very lowPeriphrastic DO constructions are very common both in English and in the Neo-Brittonic languages and are used for various functional purposes. These form part of a larger linguistic area in western and northern Europe. The literature does not mention comparable constructions for Irish and Scottish Gaelic. Irish informants, however, confirm orally that they are in common use among present day Gaeltacht speakers. They appear also to have been common in late spoken Manx. This study is based on the "Caint Chonamara" electronic corpus, the field work for which was first untertaken by Hans Hartmann (Hamburg) and Tomás de Bhaldraithe (Dublin) in the early 1960s und brought to a close by Arndt Wigger (Wuppertal) in the 1990s. The file for the Ros Muc dialogues yielded a very low return of potential DO constructions, i.e. 14 tokens of DÉAN + VN out of 494 DÉAN tokens altogether in the file. This shows that the DÉAN + VN construction was grammatically correct and acceptable to the native speakers, but was not grammaticalised and had a very low frequency. This result is interesting, but not surprising, since the informants chosen for this file conformed to the NORMS category (non-mobile old rural males. They were born around the turn of the 19c/20c and acquired their language now more than 100 years ago. This was well before the independence of the Republic. They would have acquired their Irish orally from native speakers and underwent very little formal training in Irish, or none. This small sample confirms that Irish did not belong to the broad linguistic area in Western Europe which makes use of periphrastic DO constructions, at least not until very recently.…