The final lecture of the IAS Invited Lecture Series in Multilingual History of English, organised by the Faculty of English and convened by professors Matylda Włodarczyk and Marcin Krygier, will take place on June 10, 2022 at 3.00 p.m. CEST.
Prof. Jukka Tyrkkö (Linneus University)
“Quantitative and computational perspectives to historical multilingualism studies”
Abstract: Following many decades of research focusing almost exclusively on monolingual language use, multilingualism has recently become one of the hot topics in linguistics (see, e.g., Gardner-Chloros 2009, Matras 2009, Sebba 2011). Whilst contacts with, and influences of, different languages have perhaps always been more naturally in focus in historical linguistics (e.g. Skaffari 2018), here, too, multilingualism research has gained new ground (Schendl & Wright Eds. 2011, Pahta et al. 2018), perhaps particularly when it comes to the study of code-switching as a linguistic practice and strategy, rather than as a phenomenon associated with lacking proficiency or as a feature of an ongoing process of language contact. Another key area of recent innovation in historical multilingualism has been the increasing use of corpus methods, which can help us unlock variationist patterns of usage in new ways and in increasingly large datasets (Pahta & Nurmi 2006, Nurmi et al. Eds. 2017, Demo 2018)
Accordingly, this paper will discuss the corpus linguistic side of multilingualism studies with particular reference to historical data. Reflecting on two recent research projects, one dealing with code-switching in Late Modern English (Nurmi et al. 2018, Tyrkkö & Nurmi 2017) and the other more specifically in Punch magazine (Tyrkkö & Legutko to appear), I will discuss methods of identifying, retrieving and classifying code-switches in historical data, including a demonstration of a custom corpus tool designed for this purpose (Tyrkkö et al. 2017). Next, we will look at some quantitative and statistical approaches to analysing corpus evidence of code-switching as a sociolinguistic phenomenon, focusing in particular on requirements regarding the corpus itself, as well as some general statistical issues to do with analysing phenomena of this type. Select findings from the two projects will be used to exemplify the approach and to highlight some specialised phenomena, such as code-switching to non-standard Englishes (Beal 2006). Finally, the paper will address the current and future role of machine learning in language identification and code-switching research, commenting on both the promises and challenges in this area.
The lecture will be held on Zoom. Please register at https://forms.office.com/r/nNExCj6Mrn to receive the link to the online meeting.