This thesis investigates the standardisation of Modern Scottish Gaelic orthography from the mid-eighteenth century to the twenty-first. It presents the results of the first corpus-based analysis of Modern Scottish Gaelic orthographic development combined with an analytic approach that places orthographic choices in their sociolinguistic context. The theoretical framework behind the analysis centres on discussion of how the language ideologies of the phonographic ideal, historicism, autonomy, vernacularism and the ideology of the standard itself have shaped orthographic conventions and debates. It argues that current spelling norms reflect an orthography that is the result of compromise, historical factors and pragmatic function. The research uses a digital corpus to examine how three particular features have been used over time: the dialect variation between < eu > and < ia >; variation in s + stop consonant clusters (sd/st, sg/sc, sb/sp); and the use of the grave and acute accents. Evidence is drawn from the Corpas na Gàidhlig electronic corpus created at the University of Glasgow: the sub-corpus used in this study includes 117 published texts representing a period of over 250 years from 1750 to 2007, and a total size of over four and a quarter million words. The results confirm a key period of reform between 1750 and the early nineteenth century, and thereafter a settled norm being established in the early nineteenth century. Since then, some variation has been acceptable although changes and reform of some features have centred on increasing uniformity and regularisation.

This thesis explores the measurement of adult second language (L2) oral proficiency in Scottish Gaelic (henceforth Gaelic). Gaelic is a minority language in Scotland, and is currently the object of a major effort to reverse language shift. Adult L2 users of Gaelic have been identified as key agents in this effort, but some weaknesses in adult Gaelic language-in-education policy are making it difficult for adult L2 users to fulfil this role. One such weakness is the absence of an empirically-derived means of assessing proficiency in Gaelic, through which adult L2 users and their teachers can assess their progress. This project aims to address this weakness. Data from two tasks — an interview and a narrative — performed by adult L2 users of Gaelic are analysed from the perspective of the complexity, accuracy, and fluency framework, as the three main dimensions of proficiency. Data are also analysed for Communicative Adequacy, using raters’ judgements. These data provide the first examination of Gaelic L2 proficiency from the perspective of second language acquisition (SLA) research. Adult L2 users of Gaelic have a wide range of learning experiences and motivations for learning the language. This study also explores these experiences and motivations, and discusses how these relate to proficiency. Results show that individuals’ Gaelic language skills interact in complex and unpredictable ways, depending on the nature of the task being performed. There is some evidence that the interview task encourages complexity and fluency, while the narrative task encourages accuracy at the expense of complexity. Results also show that the Communicative Adequacy rating scale developed for this project is valid and reliable, but that assessments of proficiency are subjective, to a large extent. Finally, the results confirm that adult L2 users of Gaelic draw on a vast range of experiences and are motivated in many different ways to learn the language. The outcomes of the project contribute to existing scholarship on the experiences and motivations of adult L2 users of Gaelic, confirming previous findings. The results also confirm previous findings in second language acquisition research that complexity, accuracy, fluency, and Communicative Adequacy in an L2 interact in complex ways, and that these interactions can be mediated by different task conditions. Finally, the outcomes of this exploratory research serve as the basis for future, more large-scale research into the acquisition of Gaelic as a second language by adults.

Tha an tràchdas seo a’ coimhead air buidheann de dhàin fhada a nochd aig toiseach an fhicheadaimh linn. Cha deach cus ùidh a shealltainn sna dàin seo bho sgrùdairean Gàidhealachadh sa chiad dol a-mach. Bho chrìoch an fhicheadamh linn, tha ùidh air nochdadh sna dàin seo a-rithist, gu h-àraidh An Cuilithionn, Aeòlus agus Balg agus Mochtàr is Dùghall. Ach ’s e an tràchdas seo a’ chiad sgrùdadh domhain air a’ bhuidheann seo de dhàin nan aonar. Gus an rannsachadh seo a dhèanamh thèid corpas a lorg am measg leabhraichean agus irisean bhon àm, agus na co-chruinneachaidhean a chaidh fhoillseachadh nas anmoiche. Thèid taghadh a dhèanamh a tha a’ riochdachadh prìomh fheartan nan dàn seo. Thèid rannsachadh a dhèanamh air cuid de dh’eisimpleirean bhon 18mh agus 19mh linn ann an Caibideil 2. Air sgàth ’s an uidhir de shurbhaidhean nas fharsainge a tha ri fhaotainn de litreachas na Gàidhlig, seallaidh an sgrùdadh sin air feartan sònraichte. Chìthear mar a chleachd bàrdachd Mhic Mhaighstir Alasdair cruth a’ chiùil mhòir gus gnè-sònraichte ùr de dhàin fhada a chur air bog ann an 1751. Ghabhar sùil mar an ceudna air Uilleam MacDhùn-Lèibhe air tàilibh a bhuaidh air Somhairle MacGill-Eain, agus gu robh amasan nàiseantach agus socio-cànanachais an Ìlich ri fhaicinn anns a’ mhòr-chuid de bhàird a’ chorpais cuideachd. Seallaidh Caibideil 3 air na bàird fhèin, gu h-àraidh na beachdan aca agus na ceanglaichean eatorra. Bha iad seo ri fhaicinn ann an saoghal litreachais agus poilitigs. ’S e prìomh amas na h-anailisean (Caibideilean 4-10) tuigse nas fheàrr fhaighinn an dà chuid air na dàin fhèin agus air ceistean teòiriceil a tha a’ buntainn ris a’ chorpas. ’S e an dàrna amas barrachd dhòighean-obrach teòiriceil a chleachdadh sa Ghàidhlig gus bruidhinn air litreachas na cànaine. Stèidhicheadh na h-anailisean air na feartan a bha ri fhaicinn anns gach dàn, a leithid iar-phlanntachais ann am Mochtàr is Dùghall, air neo air coimeasan a bha gus tuigse nas fheàrr a thoirt seachad air a’ chorpas shlàn. Bidh na h-anailisean seo a’ dearbhadh cuid de na ceanglaichean cudromach a th’ aig litreachas na Gàidhlig ri litreachasan Eòrpach eile, air uairean airson a’ chiad uair. Tha an co-dhùnadh a’ tilleadh do sealladh nas fharsainge air a’ chorpas. Chìthear gu bheil buidheann sònraichte de dhàin ri fhaicinn bhon chiad leth den fhicheadamh linn, dàin nach deach a sgrùdadh còmhla riamh. Tha na dàin seo ceangailte ann am meud, cuspairean, iomraidhean agus amasan nam bàrd. Leasachaidh tuigse nas fheàrr air na dàin seo an t-eòlas a th’ againn air bàrdachd na Gàidhlig, gu h-àraidh sna bliadhnaichean nuair a bha na nua-bhàird a’ nochdadh an toiseach.

The study of place-names containing personal names is a neglected field in onomastics, despite being of great significance in various areas of name-studies. At its core, this thesis will begin to bridge the gap between the study of place-names and personal names, both practically and theoretically. The first step is to introduce a formally accepted terminology for the study of these names. Here, the term used to describe a place-name containing a personal name is anthropo-toponym. The acknowledgement of such a term would aid and indeed encourage future studies of anthropo-toponyms, both in Scotland and elsewhere. The study is approached through a close investigation of name material from the Isle of Lewis. The toponyms in question are characterised by two main linguistic layers, Old Norse and Scottish Gaelic, both of which have been included here. Although this material is partially an exercise in investigating the characteristics and properties of anthropo-toponyms, it also sheds considerable light on the social and linguistic history of Lewis place-names. Additionally, the study draws on a considerable amount of comparative evidence. This is primarily collected from the comprehensive survey of The Place-Names of Fife by Simon Taylor with Gilbert Márkus (2006-12). However, when studying the Norse dimension further, material from Landnámabók, one of the key sources for the medieval settlement of Iceland, has also been included. One of the most significant proposals made in this thesis is the concept of using a variant of the name-semantic approach, previously discussed by Peder Gammeltoft (2001a) in a Scottish context. At its core, this means that rather than emphasising the etymology of individual place-name elements, the motivation for coining is emphasised. It will become evident that using this approach makes it possible to view anthropo-toponyms in a different light. Through this method, we find that there is considerable variety to be found within the name-material, particularly when we look at the social and cognitive factors at play when place-names are coined and transmitted. Place-names that, on the surface appear to be relatively homogenous, can prove to be the opposite. For example, names such as Creagan Iain Ruaidh, Geodha Bean, Mhurchaidh, Stac Dhomhnuill Chaim and Tigh Mhaoldònuich, which are all coined in a comparable social Gaelic setting in the early modern period, appear to represent motivations relating to a birth, a drowning, the abode of a notorious outlaw, and the temporary hideout of a sheep thief respectively. By emphasising these micro-narratives, it is possible to shed light on the name material from a new perspective and to provide a greater understanding of the process of coining place-names.

This thesis deals with the topic of compounding in Scottish Gaelic, in particular the relationship between marked adjective + noun compounds (e.g. deagh oidhche ‘good night’) and their unmarked noun + adjective alternatives (e.g. oidhche mhath, literally ‘night good’). The first part of the thesis surveys the relevant linguistic literature on compounding, with a specific focus on compounding in Celtic languages, and the research question for the investigation is specified as follows: in which circumstances do Gaelic speakers use the marked adjective + noun constructions in contrast to the umarked alternatives? The investigation itself consists of two experimental studies. The first of these is a corpus study using a 2.5 million word, mainly 20th century subcorpus of Corpas na Gàidhlig, a 19 million word corpus of Scottish Gaelic texts covering a range of genres, dialects and historical periods. In the course of this study, all phrases in which these adjectives occurred were collected and analysed. To annotate data the freeware AntConc concordance package was used, followed by a statistic analysis of the annotated phrases. A range of hypotheses were developed as a result of the corpus study, for example: (a) the role of conceptuality, subjectivity and emphasis in the use of preposed adjectives as opposed to their plain counterpart; (b) the religious register of ma(i)th good; and (c) the role of contrast in the use of attributive plain adjective aosta compared with the highly productive use of the preposed adjective seann- to convey the meaning ʻold’. The second experimental study involved interviews with native speakers, with the aim of providing further evidence relating to these hypotheses. 10 interviews were conducted with native speakers of Gaelic, mainly applying pictures and translations of the relevant adjectival phrases, investigating the meaning, stress pattern and hyphenation of these, as well as the default adjectives in loan words or with tangible nouns (e.g. vehicles, animals), etc. This combined methodology revealed various factors that influence the choice between marked adjective + noun and unmarked noun + adjective constructions, including: (a) dialect (with the overall use of preceding adjectives in South Uist, but the preference for math ʻgood’ in Lewis); (b) register (ma(i)th in religious texts); (c) conceptualisation in the vocabulary (by the preceding adjectives deagh- ʻgood’ and droch- ʻbad’ qualifying abstract concepts, whereas the plain adjectives math ʻgood’ and dona ʻbad’ tangible and countable nouns); (d) pragmatic factors such as the emphatic nature of deagh- as opposed to math; and (e) grammatical factors (see the use of deagh-/droch- in subjunctive clauses, particularly in time expressions).

Tha an tràchdas seo na eachdraidh chultarach air Ìle, eilean ann an ceann an iar-dheas na Gàidhealtachd, le cuideam air 1890 gu 1960. Tha e ag amas air cultar Gàidhlig nan Ìleach a chur na cho-theacs sòisealta, le bhith a’ beachdachadh air an t-suidheachadh anns an nochdadh cuid de na gnèithean cultarach a bu chudromaiche, agus air na daoine a bha an sàs ann a bhith gan cumail suas. Tha e airson an dàimh eadar cultar agus an comann-sòisealta nas fharsainge a sgrùdadh, le a bhith a’ smaoineachadh air cultar mar dhòigh air coimhead air an t-saoghal, agus mar phàirt bhunaiteach den t-saoghal sin. Tha obair a’ Mharxaich Eadailtich, Antonio Gramsci, a’ toirt buaidh mhòr air an rannsachadh, agus ’s ann air ceistean a thaobh cumhachd, clas agus gnè a thathas gu tric a’ cnuasachadh. ’S e prìomh amas na h-obrach atharrachadh taobh a-staigh a’ chultair a thuigsinn anns an àm-sgrùdaidh shònraichte seo, nuair a b’ urrainnear fhathast bruidhinn air ‘coimhearsnachd Ghàidhlig’ ach far an robh a’ Bheurla a’ sìor fhaighinn làmh an uachdair. ’S ann air trì prìomh chuspairean a tha an rannsachadh: beul-aithris, bàrdachd agus creideamh. Tha caibideil air cruinneachadh beul-aithris Eachainn MhicGillEathain agus Iain Òg Ìle (1859-1862) rud beag ron phrìomh àm-sgrùdaidh, agus tha seo a’ coimhead air obair a’ chruinneachaidh fhèin, àite nan seann sgeulachdan ann an Ìle, agus air na sgeulaichean Ìleach. Thathas an uair sin a’ gluasad gu cruinneachadh beul-aithris Elizabeth Kerr a bha stèidhichte ann am Port Sioba, san Roinn Ìlich, eadar mu 1893 agus 1910, agus a’ sireadh fiosrachadh air cuimse eadar-dhealaichte a’ chruinneachaidh seo agus na tha toraidhean na beul-aithris ag innse mun chomann-shòisealta. Anns a’ chaibideil air bàrdachd san 20mh linn, thathas a’ faighinn a-mach mun cheathrar bhàrd mu dheireadh a bha foillsichte, air na bha aca ann an cumantas agus àite romansachais agus impirileis nan cuid obrach. Às dèidh seo, tha caibideil air òrain anns na clàraidhean-fuaim Ìleach a rinn Sgoil Eòlais na h-Alba eadar 1953 agus 1969. Thathas a’ beachdachadh air na fiosraichean, air na seòrsaichean òran a bha rim faotainn – agus nach robh – agus làthaireachd òrain nam bàrd Ìleach foillisichte. Bithear cuideachd a’ toirt cunntas air buaidh mhòr a’ Chomuinn Ghàidhealaich anns an eilean. Mu dheireadh, anns a’ chaibideil air creideamh tha ceist oirnn mu dè cho eadar-dhealaichte ’s a bha suidheachadh a’ chreideimh ann an Ìle an taca ri ceàrnaidhean Gàidhealach eile, thathas a’ cleachdadh staitistig airson obrachadh a-mach dè cho cudromach ’s a bha gach buidheann-eaglais, a bharrachd air fianais beul-eachdraidh airson smaoineachadh air àite a’ chreideimh ann am beatha dhaoine.

Finn mac Cumaill (Fionn Mac Cumhaill) has always been a popular figure in Gaelic tradition, coming to full prominence during the Early Modern period, as Fenian stories (tales of Finn and his fían, or fianna, known as fianaigecht in Old Irish and fiannaíocht in Modern Irish) become ever more popular in manuscript form. Despite the popularity that both Finn and the Finn Cycle have enjoyed in Gaelic literature, mentions of Finn's death are scant and tales recounting the event are even rarer. In the extant medieval Irish literature, the pinnacle of the corpus, Acallam na Senórach, not only holds the events in relative obscurity but its presentation of the circumstances of Finn's death may even be said to be conflicting. In looking at other tales in the fíanaigecht corpus, while we find a number of references to the fact than Finn dies, only a few depict his demise, namely Aided Finn and Tesmolta Cormaic ocus Aided Finn. To this short list of narratives detailing Finn's death and the events preceding it, we can add the tale designated 'The Chase of Síd na mBan Finn and the Death of Finn' (henceforth 'The Chase') preserved in a single manuscript, London, British Library, MS Egerton 1782. Although the tale breaks off with Finn still alive, albeit weary and bloodied and standing alone encircled by his adversaries, his death is a logical next element in the narrative, not least because there is repeated mention of a prophecy of his demise throughout the tale. This tale, which spans eight manuscript pages, seems to be the longest engagement with the idea of Finn's death in the medieval and Early Modern Irish corpora, yet has been the subject of very little scholarly investigation to date. This regrettable lacuna in scholarship on Fenian literature is the starting point for this thesis, which presents a three-pronged investigation of 'The Chase'. Following a fuller introduction to the topic in Chapter 1, the history of the manuscript is examined afresh in Chapter 2 as new evidence, particularly from the works of the scribe Muiris Ó Gormáin, has shed new light on the manuscript's history and on the tale of 'The Chase'. This is then employed to examine the section of the manuscript in which 'The Chase' is to be found, a section consisting of four tales thought to be from the now lost manuscript, Cín Dromma Snechtai, and four fíanaigecht tales. It is investigated if the unit may be considered a deliberate anthology and whether thematic and/or other concerns motivated the unit's compilation. Next, the study turns to the tale of 'The Chase' itself, examining its place within a continuum of traditions found in Old, Middle and Early Modern Irish treatments of Finn's death. Based on my own linguistic work on 'The Chase', a semi-diplomatic edition of which is included as an appendix to this thesis, it is demonstrated in Chapter 3 that the author of 'The Chase' seems to have been aware of several accounts of Finn's death, either those which are now extant or sources akin to them, and sought to bring together many of the elements present in other accounts of Finn's death in a single tale, perhaps in what was intended to be a comprehensive death tale for Finn. The various elements of the tale which resonate with the event of or events leading up to Finn's death, however, have not merely been cobbled together. Rather it is illustrated that the composition skilfully treats of the themes of death, prophecy and youth versus age, making regular allusion to the audience's presumed knowledge of other tales of the Fenian corpus, while adhering to the norms of earlier written fíanaigecht literature, a trait not always found in Early Modern tales of the Finn Cycle. The last study which forms part of this thesis, Chapter 4, arose from the recognition that although 'The Chase' appears to be the longest extant engagement with Finn's death, there exists no study that details what material on Finn's death has circulated in the modern period. This section provides a comprehensive overview of modern engagements with Finn's death in post-1650 manuscripts and folklore collections. All the modern accounts that I have found to date in which Finn's death is recounted or in which it is presumed that Finn is dead, which are usually mentions of Finn's grave, are therefore identified, presented, and where applicable, translated. While it becomes clear that no other engagement with Finn's demise across the eleven centuries during which his death excited the Gaelic imagination is as long or as complex as 'The Chase', common or notable motifs in the modern accounts are identified, and similarities between the different treatments of Finn's death in the modern narratives are discussed. It is shown that a small number of the motifs and events treated in the medieval accounts of Finn’s death and in 'The Chase' are also treated in the modern tales of his demise, thus indicating some thematic continuity between medieval and modern approaches to relating how Finn died. With this in mind, some further relationships between the modern accounts of Finn's death and other medieval and modern Fenian literature are explored.