UMLV Summer School 2023

This workshop will focus on the ways in which speech can sound different when encountered in daily life and the way the brain deals with this variation. This workshop will start by defining ‚variation in speech‘ and discuss behavioural and cognitive neuroscience methods for studying how the brain processes variation during speech perception in three sessions. In Session 1, I will discuss different ways in which processing variation in speech can be studied using behavioural paradigms inspired by cognitive psychology, including response time tasks, pupillometry, and eye-tracking. In Session 2, I will start by providing a short introduction to neuroanatomy. Next I discuss four neuroimaging (fMRI, functional, Magnetic Resonance Imaging) studies that apply the paradigms and concepts from Session 1 in to neuroimaging setup. In the final session, we will run a journal club and you will get a chance to critically analyse an fMRI study investigating perception of speech in second language (L2) listeners. This workshop will make use of Mentimeter for audience response, to hear your thoughts, conduct some short quizzes, and get some general feedback.

Before the workshop, I suggest you read the paper to be discussed in the Journal Club in the 3rd session. You will have some time to read the paper in the session, but it would be good if you have had a basic understanding of the study before the session. The paper will be provided, here’s the reference: Rammell, C. S., Cheng, H., Pisoni, D. B., & Newman, S. D. (2019). L2 speech perception in noise: An fMRI study of advanced Spanish learners. Brain Research, 1720, 146316. https://doi.org/10.1016/j.brainres.2019.146316.

Finally, for those without a background in neuroscience or neuroimaging, I have added some background notes to some slides. I also recommend reading some chapters in this book: https://www.taylorfrancis.com/books/mono/10.4324/9781351035187/student-guide-cognitive-neuroscience-jamie-ward.

A discussion round on one of Dr. Theresa Biberauers recent papers will be led by Prof. Doreen Georgi (University of Potsdam), the head of the CRC 1287 on Limits of Variability in Language.

Biberauer, T. (2019). Factors 2 and 3: Towards a principled approach. Catalan Journal of Linguistics (Special Issue), pages 45-88.

The paper is strongly related to what would have been the topic of the planned workshop by Dr. Theresa Biberauer. The session will give you the opportunity to engage with the topic to some capacity and have an open discussion about it. We highly encourage you to read the paper beforehand to facilitate a more lively exchange. Dr. Biberauer also kindly offered to be available for discussions to anyone interested in the topic via email.

PCIbex (https://farm.pcibex.net/) is a platform for online behavioral experiments. It provides a simple and accessible experiment design interface, and allows sharing experiments via web browser for both data collection and Open Science resource sharing. It uses its own mini-language, which is set up to be maximally accessible while providing full control over what happens when (and where) in a given trial. The central ingredients are elements (text, images, audio, video, timers…) which can be subjected to various actions (presentation/playback…) with detailed control over timing and location. It can incorporate a diverse range of simple and complex experimental tasks with dynamic and interactive features (e.g., visual stimuli, dynamically unfolding trial structure, response feedback, scripted/ timed events, playing audio and video), and also integrates the original functionalities of its predecessor IBEX’s (Drummond 2007), including self-paced reading and rating studies. Furthermore, PCIbex offers several advanced functionalities, including audio and video recording, and an integration of the webgazer eyetracking API (Papoutsaki et al., 2016; see Slim et al. 2022 for PCIbex implementation). While these require additional initial storage-setup on the experimenter’s own server space, these capacities substantially lower the bar for deploying more sophisticated paradigms in online experimentation.

The tutorial aims to introduce PCIbex to a wide audience, with no pre-requisites or prior technical skills required, but will also be of interest to researchers with more background. The first session walks through the creation of a trial in a sample experiment combining display of un-folding text, images, and audio with different response input options. Participants can directly follow along to learn the basic syntax of the PCIbex mini-language, as well as the logic of basic trial structure, including audio and visual resources. The second session will be entirely dedicated to exercises and hands on practice to deepen participants’ understanding of working with the platform. The third session illustrates the creation of a multi-trial experiment using a template, and also goes over the integration with recruitment platforms and other practical matters of actually running experiments using PCIbex. The last part of this session will provide a brief overview and pointers towards using more advanced functionalities. Overall, the workshop aims to provide participants with a solid basis for subsequently using and exploring the platform independently to develop their own experimental projects with help from the extensive online resources available in the PCIbex documentation (https://doc.pcibex.net/).

Note: In order to actively participate and engage with the platform throughout the workshop, participants should bring a laptop with Chrome or Firefox browser installed.

For a more extensive overview of the platform and a previous similar (though shorter) tutorial, see:

Schwarz, Florian & Jeremy Zehr. 2021. Tutorial: Introduction to PCIbex – An Open-Science Platform for Online Experiments: Design, Data-Collection and Code-Sharing. Proceedings of the Annual Meeting of the Cognitive Science Society 43(43).

Slim, M.S., Hartsuiker, R.J. Moving visual world experiments online? A web-based replication of Dijkgraaf, Hartsuiker, and Duyck (2017) using PCIbex and WebGazer.js. Behav Res (2022). https://doi-org.proxy.library.upenn.edu/10.3758/s13428-022-01989-z

Deep learning methods have gained more and more relevance in various fields including psycholinguistics. In this workshop, participants will have the chance to learn about state-of-the-art deep neural models that can be used to imitate human reading behavior by generating scanpaths. The sessions will introduce all the necessary concepts from machine learning research on beginner-level and show potential use cases of such methods for research in psycholinguistics.

Since participants will complete online exercises, they are asked to (i) bring their laptops and (ii) set up a Google account before the workshop starts. Python exercises will be done in Google Colab so that no instalment of Python is required.”

Experimental methods of linguistics afford us to better understand human cognitive processes both in health and in speech-language impairments. This workshop would focus on understanding linguistic variability by studying individuals who present language disorders following neurological impairments. I would discuss linguistic variability by bringing perspectives from two lines of investigations:

– Speech and language impairments following acquired neurological disorders (e.g., post-stroke aphasia, dementias)

– Bilingualism in acquired language disorders, especially from under-researched linguistics communities.

I would conduct the workshop through three interrelated sessions aimed at:

1. Understanding linguistic breakdown following neurological impairments, such as aphasia, dementia, apraxia of speech

2. Exploring lexical level deficits in monolingual and bilingual individuals with emphasis on disentangling linguistic and cognitive influences

3. Understanding the importance of moving beyond single word investigation to connected speech in exploring the symptomatology in post-stroke aphasia and Alzheimer’s Disease for better diagnosis and assessments.

I would bring knowledge and methodological expertise from clinical linguistics, neuropsychology, speech-language therapy, neurology and psychology. In addition to learning theoretical background on the above-mentioned topics, participants would get hands on experience in analysing linguistic data from acquired neurological disorders and learn about interpretation of these data. They would also appreciate the potential of using simple linguistic tasks to important theoretical questions about the nature of linguistic variability in monolinguals and bilingual populations. Through these sessions, I would also highlight the of importance of studying under-explored linguistically diverse communities for better theoretical and clinical understanding of linguistic diversity.

This workshop will lay out how Construction Grammar can be applied to the study of language change. We will talk about the theoretical foundations, open questions, and methodological approaches that inform the constructional analysis of diachronic processes in language. The four sessions address issues such as constructionalization and constructional change, constructional networks, shifts in collocational preferences, and differentiation and attraction in the constructional network. These processes will be illustrated on the basis of studies that utilize modern corpus-linguistic methodologies and that draw on current theoretical discussions in usage-based linguistics.

No prior knowledge of Construction Grammar or corpus-based methods is required. Participants who have time to prepare are invited to download the book „Ten Lectures on Diachronic Construction Grammar“ (open access: https://brill.com/display/title/56854) and read the first two chapters.