Plenary Speaker Information

Abstract: Metaphors are abundant in daily language. What are the cognitive-neural mechanisms underlying the comprehension and production of metaphors, and what are some of the important functions of metaphors? In this talk, I will first discuss the fundamental aspects of my research that show that temporal language is grounded in motion-related areas in the brain and that metaphoric actions leverage physiological responses implicated in literal-concrete actions. Then, I will discuss the applied aspects of my research that show that the use of metaphors in abstract concept learning can improve learning in first language users and impair learning in second language learners. Specifically, after metaphoric reasoning, abstract concepts show enriched changes in the neural representations of those concepts. At the end of the talk, I will discuss limitations and future directions.

Dr. Vicky T. Lai is an associate professor of Psychology and Cognitive Science at the University of Arizona. She directs the Cognitive Neuroscience of Language Laboratory (https://sites.arizona.edu/neurolanglab/), where they use behavioral, electrophysiological, and imaging measures to study figurative language, emotion and language, and language and thought. Dr. Lai is an elected fellow in the Psychonomic Society in 2018 and received the Distinguished Mentoring Award in 2023.

Abstract: If we ask social media, Conversational AI like DeepSeek, Gemini and ChatGPT can do anything: summarize our sources, analyze our data, write up our results. It is therefore not surprising that the question on many a metaphor researcher’s mind is “Can’t I just ask ChatGPT to identify the metaphors for me?” In this talk I will address this question by discussing some of the facts and fictions concerning the use of Large Language Models (LLMs) and Generative Artificial Intelligence (GenAI) to identify, analyze and translate metaphors in authentic discourse. I present three (ongoing) collaborative projects: a study exploring how the contextualized embeddings generated by LLMs can be combined with different machine learning approaches to serve as a flexible, adaptable semi-automated data annotation tool1; a study on prompt-based metaphor identification using GPT-4 investigating the usefulness of LLM architectures for metaphor detection and sub-type labelling2; and a study evaluating the performance of different state-of-the-art Machine Translation (MT) systems and Large Language Models (LLMs) in metaphor translation in literary texts, examining the way different types of metaphor are handled by the systems and the types of errors they make. The insights provided by these projects will be used to raise awareness in the RaAM community on both the potential and pitfalls, as well as practical and ethical implications, of using LLMs in metaphor research. When we go beyond the hype, the real question is: “When is AI a help, and when is it a hindrance?”

Dr. A.G. (Lettie) Dorst is an Associate Professor in Translation and Human-Centred AI at Leiden University Centre for Linguistics, Leiden, the Netherlands. Her research focuses on metaphor variation, metaphor translation, literary machine translation, and machine translation literacy. She recently led an NRO Comenius Senior Fellow project on "The Value of Machine Translation for the Multilingual Academic Community" and was lead linguist on the ZonMW project "Dementia in Metaphors".  She is currently leading an NWO Vidi project on “Metaphors in Machine Translation: Reactions, Responses, Repercussions” (2025-2029).

Abstract: Sensory words bridge the gap between perception and language, allowing us to express what we see, hear, feel, taste and smell. Yet not all sensory experiences are equally easy to express—while English teems with adjectives for vision (e.g., "blue", "shiny", "blurred"), it offers far fewer for smell (e.g., "musky", "fragrant"). This asymmetry raises a fundamental question: how do we communicate sensations for which our language lacks dedicated vocabulary? In this talk, I explore how metaphor and iconicity work hand in hand to fill these lexical gaps. I present findings from two new empirical studies. A meta-analysis of 38 published datasets of synesthetic metaphors—expressions such as "smooth melody" or "rough taste" that blend terms from different sensory modalities—shows that across 14 languages, consistent patterns emerge in how sensory words combine. Second, I present a corpus study of 6,000 field guide entries describing bird vocalizations, which demonstrates how ornithologists rely on a combination of metaphor and onomatopoeia/iconicity to convey auditory detail in the absence of a rich vocabulary of auditory terms. Together, these studies illustrate how metaphor renders the ineffable effable.

Dr. Bodo Winter is a Professor of Linguistics and a UKRI Future Leaders Fellow. His research uses data science-driven linguistics to study multimodal communication, including iconicity, gesture and metaphor. For his Future Leaders Fellowship, he investigates how people communicate numerical information across different communication channels.