Attitudes to the Irish language in the Irish Diaspora
Patricia Ronan (Chair of English Linguistics, TU Dortmund)
This presentation explores the research questions 1) what role the Irish language plays in identity creation in diasporic Irish communities and 2) which role the arguably increased international visibility of Irish through international media plays in attitudes towards Irish. To carry out the study, an online questionnaire was compiled in Lime Survey and distributed via the social media platform X (formerly Twitter) using a snowball system. The study focuses on members of the Irish diaspora who do not predominantly identify as Irish speakers in order to obtain a broad view on attitudes to the Irish language.
Results show in how far the Irish language is important as an identity marker for the Irish diasporic community, which items beyond the Irish language are important for diasporic identity creation, and to what extent perceived modernity on the one hand and notions of usefulness on the other hand contribute to language attitudes.
About Patricia Ronan
Chair of English Linguistics since 2016. Her research focuses on the languages of Britain and Ireland, language contact and change, and multilingualism, with particular expertise in morphosyntactic and pragmatic variation. She combines corpus-based and sociolinguistic approaches and serves as Chair of the ICAME Board and Co-Editor of the ICAME Journal.
Automated Text Analysis for Pragmatics, Digital Humanities and Corpus Linguistics
Gerold Schneider (Titulary Professor in Computational Linguistics, University of Zurich)
Textual data can be analyzed increasingly well with automated techniques, ranging from keyword detection to Chat-GPT. Text no longer is unstructured data, but its highly structured nature can be explored and exploited by automatic methods, for many disciplines, ranging from Digital Humanities to Pragmatics.
Pragmatics, “the art of the analysis of the unsaid" (Mey 1991), seems to resist automated analysis in principle. While lexical methods are usually too surface-oriented, more recent methods dig deeper. First, the Firthian hypothesis, in which associations are learnt from context, is now accessible large-scale, with distributional semantics and word embeddings. Second, cue-based approaches (Jurafsky 2006) can pick up indirect signals. Third, world-knowledge, including cultural knowledge, of Large Language Models such as GPT-4 and 5 have increased, so that pragmatic annotation is increasingly successful. I present methods to detect speech acts automatically, using Searle’s taxonomy and the pragmatically annotated SPICE corpus (Kirk 2016) as training resource.
Indirectness is also a challenge for semantic analysis, even if to a lesser degree. I present a number of case studies, from the history of medicine to parliamentary, political and religious debates, from sentiment detection to complex ethical decisions, using a range of methods such as document classifiaction, topic modelling, word embeddings, conceptual maps and LLM prompting. I put special emphasis on evaluation: how well can a probabilistic classifier or Chat-GPT categorize texts, what can we learn from data-driven visualisations? Up to which point can we close the gap between distant and closed reading?
About Gerold Schneider
Titulary Professor in Computational Linguistics. His research spans corpus linguistics, computational models of language, digital humanities, statistical approaches, and large language models. He has published over 160 articles and several books, including Statistics for Linguists (open access) and Text Analytics for Corpus Linguistics and Digital Humanities. He leads the NLP group at LiRI and works on automated content analysis and hate-speech detection.
References
- Grimmer, Justin, Margaret E. Roberts, and Brandon M. Stewart (2022). Text as Data. Princeton University Press.
- Jurafsky, Daniel (2006). Pragmatics and computational linguistics. The handbook of pragmatics pp 578–604.
- Kirk, John (2016). The pragmatic annotation scheme of the SPICE-Ireland corpus. International Journal of Corpus Linguistics 21(3):299–322
- Mey, Jacob L. (1991). Text, context, and social control. Journal of Pragmatics 16(5):399–410
- Schneider, Gerold (2024). “Automatically detecting Directives with SPICE-Ireland”. In Martin Schweinberger and Patricia Ronan (eds.) Sociopragmatic Variation in Ireland: Using Pragmatic Variation to Construct Social Identity [Trends in Linguistics. Studies and Monographs]. Berlin: de Gruyter.
- Schneider, Gerold (2024). Text Analytics for Corpus Linguistics and Digital Humanities: Simple R scripts and Tools. Bloomsbury. https://www.bloomsbury.com/uk/text-analytics-for-corpus-linguistics-and-digital-humanities-9781350370821/