25-26 September 2025, Łódź, Poland
https://www.uni.lodz.pl/duo-colloquium
Faculty of Philology, Pomorska 171/173, 90-236 Łodź
Kathryn Batchelor
University College London
Translation Studies has never known what to do with Derrida: the difficulty and abstraction of his work mean that he is generally regarded as having little practical ‘application’ in a discipline that always has one eye on translation practice. Although there is growing interest in Derrida’s ideas at the translation-philosophy edge of the discipline, the majority of scholars and translation practitioners have little use for the deconstructionist thinker. The efforts of prominent scholars such as Rosemary Arrojo (1999) and Lawrence Venuti (2013) to reverse this state of affairs has, at best, produced assent without action.
In this seminar, I propose a new approach to exploring Derrida’s relevance for translators, focusing primarily on Derrida’s own translation practice, rather than on what he says about translation on a theoretical level. Although Derrida was the published translator of only one book in his lifetime, he frequently had recourse to translating, whether orally (in university teaching or in public lectures), or in writing (in his many other books and articles). In this paper, I examine the key features of this translation reflex, focussing on how it enables Derrida to elaborate his readings of other texts or to explore topics such as ethics and justice. I will suggest that Derrida’s uses of translation offer fresh ways of conceptualising translation, and that these conceptualisations get to the heart of what it means to be human.
Erik Angelone
TH Köln - Cologne University Of Applied Sciences
Artificial intelligence continues to gain firmer footing in the language industry, tangibly altering the working realities of language professionals. In response, many are diversifying their scope of language service provision (Angelone, 2025), working with broader arrays of assistive technologies, and collaborating on projects with varying constellations of fellow linguists. Recent socio-cognitive research rooted in the central tenets of distributed cognition (Hutchins, 2006) has shed valuable light on team interaction and performance in the language industry given these new realities, often with an eye towards identifying parameters of expertise (see Schlager & Risku, 2023; Schlager & Risku, 2024).
Given the prevalence of collaborative project workflows involving multiple linguists, this talk intends to contribute to still nascent explorations of team expertise (see Sannholm, 2021; Sannholm & Risku, 2024). Salas, Burke, and Stagl (2004) define team expertise as “social problem solving that promotes coordinated, adaptive team performance by facilitating goal definition and attainment” (p. 343), driven by a shared mental model. An expert team has “a latent understanding of the task, environment, and team and how these components interact” (Bisbey, Traylor, & Salas, 2021, p. 191). In the context of a language industry project workflow, this might take the form of a project manager, terminology manager, translation team, and editing team all being on the same page, distributing and sharing their individual expertise for the common good. But how do we know whether or not a mental model of the task at hand is truly shared and if those involved are truly on the same page, beyond having shared access to the same language assets and project specifications?
Using survey data to explore these questions, this talk will provide emic insights to team expertise in the language industry, rendering the opinions and perceptions of project managers, terminology managers, translators, and post-editors in relation to their own expertise and that of their collaborative partners. Instances of convergence and divergence in their rendered conceptualizations can help in identifying core facets of team expertise (reflected by convergence) and team friction (reflected by divergence) in language industry collaborative work. A better understanding of team expertise, as a human-centered and human-driven attribute, can contribute not only to optimizing training, but also to defining human-added value against a backdrop of augmentation and automation within a human(e) translation paradigm.
References:
Angelone, E. (2025). Diversification and portfolio careers. In C. Walker & J. Lambert (Eds.), The Routledge handbook of the translation industry (pp. 307–320). Routledge.
Bisbey, T., Traylor, A., & Salas, E. (2021). Transforming teams of experts into expert teams: Eight principles of expert team performance. Journal of Expertise, 4(2), 190–207.
Hutchins, E. (2006). The distributed cognition perspective on human interaction. In N. Enfield & C. Levinson (Eds.), Roots of human sociality: Culture, cognition and interaction (pp. 375–398). Routledge.
Salas, E., Burke, C. S., & Stagl, K. C. (2004). Developing teams and team leaders: Strategies and principles. In D. Day, S. J. Zaccharo, & S. M. Halpin (Eds.), Leader development for transforming organizations: Growing leaders for tomorrow (pp. 325–355). Erlbaum.
Sannholm, R. (2021). Translation, teamwork, and technology: The use of social and material scaffolds in the translation process (Doctoral dissertation). Stockholm University.
Sannholm, R., & Risku, H. (2024). Situated minds and distributed systems in translation: Exploring the conceptual and empirical implications. Target, 36(2), 159–183.
Schlager, D., & Risku, H. (2023). Contextualising translation expertise: Lived practice and social construction. Translation, Cognition & Behavior, 6(2), 231–252.
Schlager, D., & Risku, H. (2024). What does it take to be a good in-house translator? Constructs of expertise in the workplace. The Journal of Specialised Translation, 42, 2–19.
Miguel A. Jiménez-Crespo; Krzysztof Kowalczyk; Carme Mangiron; Agnieszka Szarkowska
Rutgers University; Polish Association Of Audiovisual Translators (STAW); Universitat Autònoma De Barcelona; University Of Warsaw
Audiovisual translation, media accessibility, and game localisation could not possibly exist without advanced technology. The aim of said technology is to embed the target text in the environment in which it will be delivered to the consumer, but also to replace the human translator in certain repetitive and mundane tasks. At the same time, a successful rendition of media content requires creativity, which is the preserve of talented and experienced human translators, and which - for now - is beyond the reach of artificial intelligence. This panel will embrace the role of technology in creative media localisation, to establish whether it is merely a vehicle for providing the final version to the target audience, whether it contributes to the translation process in much the same way as the human translator, or perhaps poses a constraint on the human translator’s visibility, authorship, and potential.
While AI can handle routine translation tasks, human translators excel at contextual accuracy, emotional nuance, and cultural authenticity. Assuming that the current state of play is collaboration of man and machine for the common good, where each party contributes their strengths to alleviate their weaknesses, the panelists will undertake to explore what it means to the translation and localisation industry at large, and which competences educators should practise with translation students. Furthermore, the question of rationale for GenAI ethics guidelines, refined copyright laws, new professional standards, rules, or regulations will be raised. How should the translation and localisation market adapt, so that consumers receive quality translations without delay, while translation agents receive credit and an optimal working environment? In other words, will humans work with or against technology?
Fear of the unknown is a common phobia. The advent of GenAI was met with curiosity, expectation, and trepidation. Surveys show that translators fear for their jobs. We may be in for a „fast food” type of future, where art, music, writing, and translation will be done by machines on a large scale - instant, superficially enjoyable, acceptable and theoretically „correct”, but bland and sorely lacking in the genius factor. GenAI is here to stay and it won’t be swept under the carpet, but does it pose a threat? Is it sufficient to educate about it, to train and control it? What steps should be taken? Perhaps it is already too late? Those are some of the questions that the panelists will tackle in this session.
Chantal Wright
School Of Applied Linguistics, ZHAW
Unlike NMT, we humans are anything but indifferent towards the literary texts that we translate. We bring our personal histories, our libraries, our bodies, our emotions and our intellects to the work of literary translation. Both the translation process and the translation product are meaningful for us. The aim of this non-language-specific workshop is to play and experiment, in order to make ourselves more visible in translated texts.
The workshop will be led in English, but participants are welcome to work with any language pair. Please bring a poem or a short prose text (an extract from a novel or a short story or from a work of literary non-fiction) that is significant for you. Perhaps the text reminds you of a particular period in your life, a relationship or a person. Perhaps it is simply a text that you love, or maybe a text that you can’t stand because you read it too often in school, perhaps it is a text that fascinates you because of its complexity or its style or its theme. If you are bringing a longer poem or a longer prose text, you can select a few lines upon which to concentrate in the workshop.
Read the text a few days before the workshop, carry it around in mind and body during your daily activities and observe your intellectual engagement with the text, your emotions, memories and any other reactions that the text provokes.
The text can be from your mother tongue/L1, your L2, or another foreign language.
Monika Porwoł
State University Of Applied Sciences In Racibórz
This paper addresses the significance of translator creativity within the context of the dynamic advancement of artificial intelligence (AI) and the concept of Society 5.0. The analysis provided includes a comparative examination of translations performed by professional translators and machine translation systems, with particular emphasis placed on human creative competencies as well as the capabilities and limitations of contemporary translation technologies. Additionally, I present the evolution of translation tools, highlighting the growing role of AI while simultaneously underscoring the irreplaceable nature of human creativity, linguistic intuition, and cultural empathy. In this study, I introduce (currently under discussion) translation paradigm based on synergy between humans and technology, indicating the necessity to redefine the professional competencies of translators in an era of advanced technologies.
Joanna Mirek (Online/Video)
John Paul II Catholic University Of Lublin
With the growing popularity of artificial intelligence in both educational and professional contexts, it is crucial to consider the potential of digital tools that may effectively support situated interpreter training.
This paper presents preliminary findings from a pilot study involving 27 second-year MA students of English Philology at the John Paul II Catholic University of Lublin participating in a one-term simultaneous interpreting course. As part of mock conference practice, the trainees were instructed to use ChatGPT to conduct background research, predict speech content, and extract terminology to prepare glossaries for mock conference speeches based on the abstract provided by the trainer. The output generated by the system was evaluated with regard to its accuracy and usability in simultaneous interpreting practice.
Drawing on the theoretical frameworks of situated learning (González-Davies, Fernández-Prieto, & Sempere-Linares), socio-constructivism, and emergent learning (Kiraly), the pilot study investigates how interpreting trainees interact with AI during the preparation stage of mock conferences. Specifically, it explores the quality of information gathering (background research), prediction of speech content, machine translation, and glossary creation.
The preliminary findings indicate that despite AI accelerating the preparation process, the output oftentimes requires correction due to terminology redundancy, inconsistency, or insufficient prediction of the speech content. Furthermore, the most demanding aspect of the task for the interpreting trainees was the in-class collaborative refining of the output, which required critical reflection.
The paper advocates for a pedagogical approach that encourages interpreting trainees to engage critically with AI, emphasising that technology should complement rather than replace the situated learning process. The findings suggest that interpreter training needs to remain human(e): both efficient and grounded in the premises of situated learning and socio-constructivism.
References:
Fernández-Prieto, C., & Sempere-Linares, F. (2010). Shifting from translation competence to translator competence: Can constructivism help? In V. Pellatt, K. Griffiths, & S.-C. Wu (Eds.), Teaching and testing interpreting and translating (pp. 131–147). Peter Lang.
González-Davies, M. (2004). Multiple voices in the translation classroom: Activities, tasks and projects. John Benjamins.
Kiraly, D. (2000). A social constructivist approach to translator education: Empowerment from theory to practice. St. Jerome.
Kiraly, D., & Hofmann, S. (2019). Towards an emergentist curriculum development model for translator education. In G. Massey & D. Kiraly (Eds.), Towards authentic experiential learning in translator education (2nd ed., pp. 59–81). Cambridge Scholars Publishing.
Li, X. (2015). Mock conference as a situated learning activity in interpreter training: A case study of its design and effect as perceived by trainee interpreters. The Interpreter and Translator Trainer, 9(3), 323–341. https://doi.org/10.1080/1750399X.2015.1100399
Katarzyna Popek-Bernat
Uniwersity Of Warsaw
The paper aims to explore the effectiveness of selected AI tools in translating legal texts between Polish and Spanish. After referring to the regulations concerning the usage of MT in professional translations in the European market and discussing some ethical issues related to that topic, the results of an experiment consisting in the translation of the fragments of publicly available judgments and sentences by TextCortex, ChatGPT, GoogleTranslate, DeepL, Bing Microsoft Translator will be presented. The translatory solutions generated by each of the AI tools mentioned above will be compared with the translations of the same texts made by professional human translators associated with the Polish Society of Sworn and Specialized Translators PT TEPIS as a result of a brainstorm session during the XXXIII Legal and Specialist Translation Workshop held in Warsaw in October 2019. Their relevance and appropriateness to the context will be assessed by determining the level of the ‘textual fit’ (Chesterman 2004; Biel 2014) of AI- and human-translated texts with non-translated texts of the same genre written in target languages of our interest (i.e. Polish and Spanish). In order to do that, the translations will be contrasted against the corpus of randomly selected rulings extracted from the public databases provided by the Judicial Documentation Centre of the General Council of the Judiciary of Spain (in Spanish: Centro de Documentación Judicial del Consejo General del Poder Judicial) and of the Common Court Ruling Library of Poland (in Polish: Portal Orzeczeń Sądów Powszechnych). Taking as a starting point the contributions of other studies to the thematic field of legal translation between Polish and Spanish (cf. Gołębiewska 2019; Urbańczyk 2022), the main focus will be put on the characterization of specific stylistic features of non-translated legal texts (e.g. the use of impersonal imperative forms, of the so-called “pasiva refleja” or of the binomial expressions typical of Spanish judicial language or the use of some periphrastic modal constructions with defective verbs and third-person singular verb forms to express a judge’s decision typical of Polish judicial language) and their detection in human vs. AI-based translations.
References:
Biel, Ł. (2014). Lost in the Eurofog: The textual fit of translated law. Peter Lang.
Chesterman, A. (2004). Hypotheses about translation universals. In G. Hansen, K. Malmkjær, & D. Gile (Eds.), Claims, changes and challenges in translation studies: Selected contributions from the EST Congress, Copenhagen 200 (pp. 1–13). John Benjamins.
Gołębiewska, M. (2019). Traducción de las resoluciones judiciales del español al polaco. Análisis de algunos problemas relacionados. In A. Jackiewicz, K. Popek-Bernat, & E. Waluch de la Torre (Eds.), Traducción al polaco de textos jurídicos españoles. Problemas, retos, soluciones (pp. 13–96). Instytut Studiów Iberyjskich i Iberoamerykańskich & Muzeum Historii Polskiego Ruchu Ludowego.
Popek-Bernat, K. (2024). Pluricentryzm języka hiszpańskojęzycznych wyroków karnych na przykładzie wariantywności w obrębie najczęściej stosowanych wyrażeń przyimkowych. Lingua Legis, 32, 27–45. https://lingualegis.ils.uw.edu.pl/index.php/lingualegis/article/view/87/74
Urbańczyk, E. (2022). “Que estimando la demanda (…), debo condenar y condeno”: Problemática del gerundio en la traducción jurídica del español al polaco. In K. Popek-Bernat (Ed.), Normatividad, equivalencia y calidad en la traducción e interpretación de lenguas ibéricas (pp. 119–148). Peter Lang.
Elena Errico, Gabriel González Núñez, Leticia Valdez Gutiérrez
Università Di Trieste (Italy), The University Of Texas Rio Grande Valley (USA), Universidad Autónoma De Baja California (Mexico)
Translation students bring into the classroom different bilingual profiles. These varying profiles may range from early bilinguals to late bilinguals. Whatever the case may be, such bilinguals often have previous experiences as language brokers (i.e., as informal language mediators for others). This experience, like their linguistic profile, may potentially affect the translation competence they bring into the classroom and their ability to develop the necessary skills for success. Understanding this, the present paper seeks to answer the question as to whether there exists a relationship between the linguistic experiences and backgrounds of bilingual students on the one hand and their development of translation competence (i.e., the necessary skills to perform at the level of professional translators) on the other. To this end, the present paper will describe and analyze a study carried out among translation students at the university level in the Rio Grande Valley, Texas, and the Tijuana municipality, Baja California.
Krzysztof W. Hejduk
University Of Lodz
Video games are a unique medium operating through interactive multimodal software that centres around the user as the indispensable actant in the process of simultaneously constructing and consuming the emerging gameplay. Therefore, understanding how players experience a given title tends to play a crucial role during its development and after its release. This is also in part why game localisation complementarily tends to assume a more functionalist approach to product quality (cf. Reiß et al. 2015), i.e. offering each locale (playerbase) an equally potent, though not necessarily symmetrical, palette of appropriate affective experiences to be elicited through gameplay – in close relation to the locale’s language, customs, political climate, etc.
With that in mind, it is standard practice in the gaming industry to test commercial games by employing specialised game testers and involving ordinary consumers for feedback. It is unknown, however, how much of this practice considers international users. The data resulting from such tests is in itself often classified, and what is mostly accessible is anecdotal or corporately devised reports of similar procedures occurring on the scale of global players and with localised versions of games. Even without directly accessing this data, the user-centric research into game translation is growing (e.g. Deckert et al. 2024: 1–13), although remains overshadowed by how prevalent and impactful the focus on users is in the industry context (Mangiron 2018: 279). Gaming is contemporarily a key trans-cultural activity with worldwide implications, mediated through the lens of linguacultures, technology, and large international communities (as of 2025, 3 billion gamers across the globe). It is also very lucrative, as arguably the biggest sector of the entertainment industry.
My talk aims to answer what game translation user research is and how it could be advanced. To that end, I present a brief account of user-centric research into the localisation of video games from historical and conceptual perspectives. I also overview how this line of research connects with some neighbouring areas of inquiry, such as the study of game users (Drachen et al. 2018) or translation end-users (e.g. Suojanen 2015). My findings into the emerging prospects of furthering user-centred research into game translation include current projects, such as examples of active collaboration between academics and game developers, which allows for “access to the code of mainstream games […] [for] manipulation of the different subtitling [and similar localisatorial] parameters [that] would be most beneficial for future tests” (Mangiron 2018b: 293).
References:
Deckert, M., Hejduk, K. W., & Bernal-Merino, M. Á. (2024). Towards game translation user research. Cambridge University Press.
Drachen, A., Mirza-Babaei, P., & Nacke, L. E. (Eds.). (2018). Game user research. Oxford University Press.
Mangiron, C. (2018). Reception studies in game localisation: Taking stock. In E. Di Giovanni & Y. Gambier (Eds.), Reception studies and audiovisual translation (pp. 277–296). John Benjamins.
Reiß, K., Vermeer, H. J., Nord, C., & Dudenhöfer, M. (2015). Towards a general theory of translational action: Skopos theory explained. Routledge.
Suojanen, T., Koskinen, K., & Tuominen, T. (2015). User-centered translation. Routledge.
Marianna Kryszkowska And Marta Kajzer-Wietrzny
Adam Mickiewicz University
Our study concerns English automated translations of Polish accommodation reviews posted on Booking.com, which are an example of user-generated content (UGC) i.e. a product of users’ active participation online. Due to its characteristics, such as the use of hashtags, or idiomatic expressions, UGC poses a challenge for machine translation and sentiment analysis (Saadany et al., 2021). Online reviews constitute a special type of UGC as they not only serve as a source of information about a product or a service but also provide users’ subjective opinions (Lula et al., 2016). Given that, accurate translations of reviews are necessary for users to make informed decisions. Translation accuracy is key not only in the case of syntax and semantics, but it is also important for the translated versions to exert a similar emotional response among the target recipients as the original.
The goal of the study is to investigate how various automated translation methods (Google Translate, DeepL Translate, and GPT-3.5) render emotional content and if their translations of emotion words differ significantly. Certain emotions lack direct translation equivalents between Polish and English, resulting in variations in human translations of basic emotion terms (Bąk 2023). Consequently, this study also aims to determine whether differences in automated translation outputs are noticeable for emotions that have fewer direct equivalents in human translations within the same language pair. The last goal is to find if the sentiment score can be predicted by the rating and the automated translation method.
To answer the questions, correspondence analysis and a linear mixed-effects model were fit. Correspondence analysis reveals clear distinctions between AI-generated translations and machine translations. Regression modelling confirms that GPT-3.5's sentiment scores align more closely with original review ratings, while DeepL Translate and Google Translate exhibit flatter sentiment curves. The results align with what has been found on human translation, meaning that positive emotions are rendered in a more consistent manner, whereas negative emotions are translated more dissimilarly. The findings point that GPT-3.5 is more sentiment-sensitive than the two other methods.
References:
Bąk, H. (2023). Issues in the translation equivalence of basic emotion terms. Ampersand, 10, Article 100128. https://doi.org/10.1016/j.amper.2023.100128
Lula, P., Wójcik, K., & Tuchowski, J. (2016). Analiza wydźwięku polskojęzycznych opinii konsumenckich ukierunkowanych na cechy produktu [Feature-based sentiment analysis of opinions in Polish]. Prace Naukowe Uniwersytetu Ekonomicznego we Wrocławiu, 427. http://www.dbc.wroc.pl/dlibra/docmetadata?id=33161&from=publication
Saadany, H., Orasan, C., Caro Quintana, R., Do Carmo, F., & Zilio, L. (2021). Challenges in translation of emotions in multilingual user-generated content: Twitter as a case study. arXiv. http://arxiv.org/abs/2106.10719
Dominik Kudła
University Of Warsaw
For several years now, the use of generative artificial intelligence (genAI) tools in translation services is a topic of intense debates among translators, scholars, as well as users of translated content. The efficiency and ethics of utilising such solutions are among the most commonly discussed issues. The question is also topical for video game localization, where linguistic means play a crucial role in creating immersive player experience. Although some providers claim that in their case genAI translation is always post-edited, game reviews and user opinions evidence otherwise.
The paper will present the results of a biometric online study conducted in the Laboratory of Media Studies at the University of Warsaw. The study featured a group of 46 Polish native speakers, demonstrating at least B2 English proficiency (tested using the Cambridge Assessment General English test) and defining themselves as dedicated gamers (confirmed by a series of gaming preferences the pre-test questions). Webcam eye tracking and face tracking was used to study the content processing and reactions of the participants watching screenshots of a Polish language version of a role-playing video game “Drakkar Crew” before watching screenshots featuring revised Polish localization of these fragments. Each participant watched six screenshots showing the originally released English-to-Polish localization of the game and six screenshots featuring the same fragments, some of which were revised by a human translator and some of which were revised using exclusively ChatGPT 4o. The biometric results were triangulated with answers provided by the participants in a questionnaire following each of the watching activities. The analysis was aimed at verifying whether the translator experience and participation in video game localization courses of the participants had impact on their perceptiveness regarding various types of errors in all the stimuli as well as successful recognition of the fragments revised using solely genAI. Although there was no significant difference found for successful recognition of genAI solutions noted between the groups, some types of faults in both the unrevised and revised localizations were discovered and correctly attributed to genAI more frequently by participants with greater translation experience or at least linguistic/ philological background. Some issues typically associated with genAI translation lacking any post-editing, e.g. use of unnatural neologisms or mistranslations affecting the logical structure of the text, evoked strong emotional reactions in most participants.
References:
Bernal-Merino, M. Á. (2015). Translation and localisation in video games: Making entertainment software global. Routledge.
Deckert, M., & Hejduk, K. (2022). On-screen language in video games: A translation perspective. Cambridge University Press.
Gackowski, T., Brylska, K., Jaśkowski, P., & Patera, M. (2018, June 17). FIFA 18 players – How do they feel and what do they see? Eye tracking, GSR/EDA, ECG and face tracking – Research reconnaissance [Video]. 6th Polish Eye Tracking Conference, Warsaw, Poland. YouTube. https://www.youtube.com/watch?v=3pA-ySbOKJ8
Kudła, D. (2024). Eye tracking in video game localization player-centric studies. In M. Deckert & K. Hejduk (Eds.), User-centric studies in game translation and accessibility (pp. 63–83). Routledge.
Moreno García, L. D., & Mangiron, C. (2024). Exploring the potential of GPT-4 as an interactive transcreation assistant in game localisation: A case study on the translation of Pokémon names. Perspectives. https://doi.org/10.1080/0907676X.2024.2378346
Lora Tamošiūnienė
Mykolas Romeris University
The concept and the presence of World Literature is much debated, the centers of cultural power are identified by multiple scholars as the linguistic and cultural presences of political powers (Damrosh, Casanova, Moretti, Beebee, Di Leo, Walkowitz). Human experience, considerations and negotiations between cultural histories, languages and individual preferences decide the slight shifts in the formation of the overall shared human understanding of the World Literature. However, a new and, potentially ,unwelcome voice in the field of literary translation, Artificial Intelligence, is attracting the foci of multiple discussions in this forming field of World Literature with numerous emerging issues.
The presentation will address the overview of reception of AI in the field of literary translation by referring to “Humane AI Manifesto 2020”, “The 2023 Manifesto on Literary Translation PEN”, CEATL supported “Manifesto for Human Language. German Joint Open Letter and Manifesto for Human Language 2024“, and Ella Creamer`s article of 8 July, 2025 in “The Guardian“ on GlobeScribe.ai service in literary translation used by several publishers.
The speed, statistical probability of linguistic construction and “good enough“ standard of literary translation production do not seem to hold water of such claims as giving the voice to people “with different ways of life and world views“, “ actively creating meanings, relationships and new language worlds, subverting familiar perspectives and conveying unfamiliar experiences“, “ shaping the work’s aesthetic effect and the “space” that authors and translators share with their readers“. AI is seen as “producing the status quo[...] multiplying biases and inhibitigs creativity.“
The presentation will also look into several samples of translations from Korea literary texts into English and Lithuanian that would provide illustrations to the acuteness of in-depth consideration of linguistic, cultural, ideological issues in creating a translated text and to the aspects these issues are changing in the status quo of the concept of the World Literature today within (a) local literature(s).
References:
Creamer, E. (2025, July 8). Review of AI translation service launched for fiction writers and publishers prompts dismay among translators. The Guardian.
Hiddleston, J., & Ouyang, W.-c. (2021). Multilingual literature as world literature. Bloomsbury.
Lukowicz, P. (2020). HumanE AI manifesto (Ref. Ares(2020)4075189 - 03/08/2020). European Commission.
Powell, A. M., Finegan, A., & Afary, F. (2023). The 2023 manifesto on literary translation. PEN America. https://pen.org/report/translation-manifesto/
Walkowitz, R. L. (2017). Born translated: The contemporary novel in an age of world literature. Columbia University Press.
Judyta Mężyk
University Of Silesia In Katowice
Music plays a vital role in shaping the emotional narrative of film, yet its accessibility often remains limited for Deaf and Hard of Hearing (DHH) audiences due to inadequate subtitling practices. Traditional closed captions (CC) frequently reduce musical elements to vague descriptors, failing to convey the emotional depth communicated through sound (Aleksandrowicz, 2020). Although streaming platforms have begun to address this gap by encouraging more detailed music captions, questions remain about the potential of artificial intelligence (AI) to contribute to this effort. Recent research indicates that AI demonstrates considerable proficiency in interpreting emotions from text, though it is still limited when faced with subtle emotional nuances without human oversight (Finet, Kristoforidis, & Laznicka, 2025). Meanwhile, in terms of emotion recognition from visual imagery, significant variability has been observed across AI systems (Elyoseph et al., 2024), suggesting that reliability in multimodal emotional interpretation remains a challenge. With this is mind, the present study investigates the effectiveness of AI-generated CC in capturing musical emotion. The analysis focuses on a Polish-language transcription produced by ChatGPT for a short animated film characterized by strong musical storytelling and no spoken dialogue. Using qualitative coding methods, each subtitle was categorized as either “neutral” or “emotional” to assess the representation of musical nuance. The output was also evaluated for compliance with widely accepted subtitling standards, including readability, timing, and semantic accuracy. These results were then compared to a human-produced CC version created by translation students (ongoing study by Rivas Ginel, Siragusa, Moorkens, & Mężyk). By examining how AI interprets and renders music-driven emotion into accessible text, this study aims to contribute to emerging discussions on inclusive media practices and the evolving role of AI in audiovisual translation and media accessibility.
References:
Aleksandrowicz, P. (2020). Can subtitles for the deaf and hard-of-hearing convey the emotions of film music? A reception study. Perspectives, 28(1), 58–72.
Elyoseph, Z., Refoua, E., Asraf, K., Lvovsky, M., Shimoni, Y., & Hadar-Shoval, D. (2024). Capacity of generative AI to interpret human emotions from visual and textual data: Pilot evaluation study. JMIR Mental Health, 11, e54369.
Finet, A., Kristoforidis, K., & Laznicka, J. (2025). The limits of AI in understanding emotions: Challenges in bridging human experience and machine perception. F1000Research, 14, 582.
Rivas Ginel, M. I., Siragusa, I., Moorkens, J., & Mężyk, J. (2024, October). Resonance beyond sound: Exploring emotions in music for the Deaf and Hard-of-Hearing with closed captions [Conference presentation]. Languages & The Media Conference, Budapest, Hungary.
Tímea Kovács
KRE, Hungary
As generative machine translation becomes increasingly capable of modelling the way natural languages work, the traditional roles of translators and others involved in the translation process are gradually being replaced by new ones (Castilho et al., 2019). Consequently, there is an increased emphasis on pre- and post-editing (proofreading) skills and competences, which facilitates the production of higher quality translations (Pym 2013; Robert et al. 2017). The efficacy of pre- and post-editing is contingent on translators’ cognisance of the mechanisms and limitations of generative machine translation tools in the pertinent language pairs, as well as the register of the relevant text. The majority of studies have found that the language of texts produced by generative translation tools is more ‘natural’ than that of earlier versions. Concurrently, they have identified recurrent omissions, redundant insertions and repetitions. Researchers have identified a lack of consistency at the text level and frequently criticise the treatment of technical terms in texts. In this presentation, the challenges encountered during the translation process will be illustrated by comparing non-human translations (generated by machine translation applications, particularly DeepL) with those produced by human translators. The present study aims to highlight the specific challenges that arise in the translation process by conducting a qualitative analysis of Hungarian legal texts, with particular reference to Act C of 2012 on the Criminal Code and Act V of 2013 on the Civil Code, as well as their English translations. The analysis will draw on both machine-generated translations and human translations, produced by translation tools and by human translators respectively.
References:
Castilho, S., Gaspari, F., Moorkens, J., et al. (2019). Editors’ foreword to the special issue on human factors in neural machine translation. Machine Translation, 33, 1–7. https://doi.org/10.1007/s10590-019-09231-y
Pym, A. (2013). Translation skill-sets in a machine-translation age. Meta, 58(3), 487–503.
Robert, I. S., Terryn, A. R., Ureel, J. J. J., & Remael, A. (2017). Conceptualising translation revision competence: A pilot study on the ‘tools and research’ subcompetence. The Journal of Specialised Translation, 28, 293–316.
Miguel A. Jiménez-Crespo
Rutgers University
“Pre-translation competence” is the set of skills and dispositions that language learners possess before they engage in advanced translation training (Presas 1998). Language learners and bilinguals also possess what can be described as “pre-translation technological competence” (Jiménez-Crespo 2017), that is, the use of translation technologies for language learning and mediation widely documented in published literature (Niño 2009, 2020; Jolley and Maimone 2022, 2025; Lee 2023). With the emergence of generative AI (GenAI) and large language models (LLMs), this pre-technological competence possessed by these language learners is now radically different. Building on the literature on translation competence, this presentation argues that although pre-translational competences may have evolved in the AI era, the development of advanced professional translation competence still requires attention to specific core components and skills that go beyond those of bilinguals or language learners. The presentation will examine the differences between bilinguals, language learners, and professional translators from the perspective of translation competence (e.g., Göpferich & Jääskeläinen, 2009; Pym & Windle, 2011b; Jiménez-Crespo, 2013; Pym, 2023, p. 101), and will discuss how these differences represent essential skills for excelling in a future shaped by human–AI collaboration. It will be shown that issues such as the use of less literal techniques and increased paraphrasing, attention to the communicative context, the ability to theoretically justify decisions within translation teams, the identification of errors and problems, and the incorporation of risk assessment, among others, not only distinguish bilinguals, novices, and professional translators, but also highlight the differences between prompted LLM translation and the skills demonstrated by professionals according to recent research.
Agata Sadza
University Of Łódź
Paulina Pietrzak
University Of Łódź
This presentation focuses on the role of translators’ personal resources in hybrid workflows. While AI tools increase efficiency, they can also foster dependency and undermine foundational skills, making it essential to balance academic rigour with market preparedness. Reflective practices help students connect experience with translation tasks, develop digital resilience and strengthen professional identity. The presentation emphasizes that education must go beyond merely adapting to AI. Students need to be empowered to shape both the translation process and themselves.
Examples of training practices presented here aim to promote a proactive approach to skill development and highlight the cultivation of key personal resources. Strategies for enhancing metacognitive translator competence are suggested as invaluable for complementing AI tools. This approach seeks to transform apprehension into empowerment, enabling future translators to control AI advancements with integrity, while remaining human, humane and purposefully reflective in a technology-driven market.
Michał Kornacki
University Of Łódź
This presentation introduces digital anxiety as a pedagogical and professional phenomenon in translator and interpreter education: the stress and uncertainty learners and educators face as AI, machine translation, automation, platformisation, and shifting industry expectations reshape tools, workflows and evaluation norms. Manifestations include fears about job security, pressure to upskill continuously, ethical dilemmas, quality-assurance worries, and work–life balance strains—often driven by systemic conditions rather than individual technical deficits.
Carme Mangiron
Universitat Autònoma De Barcelona
Abstract: Game accessibility has grown significantly in recent years. However, cognitive game accessibility remains under-researched and poses challenges due to the heterogeneity of potential users. This paper presents the WEL project (From written to oral texts in Easy Language: easy audios in cultural visits and video games, PID2022-137058NB-I00), which investigates cognitive accessibility in video games and explores whether easy language would contribute to enhancing game accessibility, focussing on players with dyslexia as a case study. First, the main cognitive barriers in video games will be presented and a number of recommendations for improving cognitive accessibility - resulting from a systematic literature review - will be outlined. Next, the results of two focus groups with players with dyslexia will be discussed. The main themes of the focus groups were players' needs and preferences and their views about the potential inclusion of easy language in video games. Finally, the next step of the project, which consists of a reception study with a video game that includes easy language will be presented.
Karolina Sawa
University Of Łódź
The landscape of Audiovisual Translation (AVT) and Media Accessibility (MA) research presents significant methodological challenges. While user experience and game playability studies continue to expand, Audiovisual Translation and Media Accessibility research focusing specifically on individuals who engage with games, whether as active players or passive viewers of gameplay, remains underdeveloped and fragmented (Deckert et al., 2024: 1). Consequently, there exists a pressing need to develop tools, resources and approaches that can propel this research area forward. This study presents a research instrument and a resource designed to address these methodological gaps: PXI-PL, a Polish adaptation of the Player Experience Inventory (PXI) based on Abeele et al. (2020), and “DISTRAINT for Research” (D4R), an online piece of software (Hejduk et al. 2025).
The Player Experience Inventory addresses a fundamental research gap regarding “how specific game design choices are experienced by players, and how these lead to specific emotional responses” (Abeele et al., 2020: 1). The PXI operates through two measurement levels: Functional Consequences, which assess how players experience game design elements such as control mechanisms and progress feedback, and Psychosocial Consequences, which evaluate emotional responses including immersion and mastery (2020: 1-10). The adapted version (PXI-PL) is currently being tested to enable comprehensive measurement of player experiences among Polish gaming communities.
The second methodological contribution emerges from Academia-Industry collaboration. D4R represents an openly available and customizable online resource specifically designed to examine how game localisation and in-game linguistic parameters influence user-game interactions (Hejduk et al., 2025). The platform was developed from Jesse Makkonen's 2D indie psychological horror adventure game DISTRAINT: Deluxe Edition (2017), which was previously used as stimulus material by Deckert and Hejduk (2022). The original experimental approach required participants to download and install PC demo versions on their personal computers (Deckert & Hejduk 2022: 64). The current D4R platform eliminates these technical barriers by providing browser-based access through URL linking to server-hosted game files, creating an ecologically appropriate research environment that minimizes hardware requirements while maintaining experimental validity. The platform's customizable architecture allows researchers to tailor the gaming experience to specific research objectives.
These complementary solutions contribute towards a comprehensive methodological framework for systematic investigation of audiovisual translation and media accessibility features in gaming contexts, providing researchers with both validated measurement capabilities and an accessible research platform.
References:
Abeele, V. V., Spiel, K., Nacke, L., Johnson, D., & Gerling, K. (2020). Development and validation of the PXI: A scale to measure player experiences at the level of functional and psychosocial consequences. International Journal of Human-Computer Studies, 135, 1–12. https://doi.org/ggjqff
Deckert, M., & Hejduk, K. W. (2024). Can video game subtitling shape player satisfaction? Perspectives: Studies in Translation Theory and Practice, 32(1), 59–75. https://doi.org/jp9m
Deckert, M., Hejduk, K. W., & Bernal-Merino, M. Á. (2024). Towards game translation user research. Cambridge University Press.
Hejduk, K. W., Makkonen, J., Sawa, K., & Deckert, M. (2025). “DISTRAINT for Research” (D4R) – A resource in game translation studies and beyond [Manuscript submitted for publication].
Makkonen, J. (2017). Distraint: Deluxe Edition [Video game]. Jesse Makkonen.
Julia Maryniak
University Of Łódź
The presentation aims to discuss the need for human translator in a scientific translation, taking Archaeology as an exemplary field.Since the emergence of generative AI, the topic of its usage in translation, especially scientific, was at a rise. This is due to two common myths that this presentation aims to disprove: (1) "AI makes the translation process quicker” and (2) "AI translation is already good enough for scientific text". To that end, the author will present examples both from the translator practice and tackle the issues of the language used in archaeology.
Archaeology is a science that verges on many others: history, architecture, biology, geography, geology and physics, to name a few. Therefore, it is characterised by an extreme versatility in the topics it tackles. Thus, the language it uses can be described as a very specific jargon delineated by a high level of abbreviation or usage of common words with new given meanings for example. Those aspects make the language of archaeology akin to a code that only one in the know can decipher and as a result translate.
References
1. Al-Smadi, H. M. (2022). Challenges in Translating Scientific Texts: Problems and Reasons. Journal of Language Teaching and Research, 13(3), 550–560. https://doi.org/10.17507/jltr.1303.11
2. Malinowski, T. (2014). O niektórych wyrażeniach w polskim piśmiennictwie archeologicznym. DOAJ (DOAJ: Directory of Open Access Journals), LIX(1-2).
3. Marek Łukasik. (2024). The Future of the Translation Profession in the Era of Artificial Intelligence. Survey Results from Polish Translators, Translation Trainers, and Students of Translation. Lublin Studies in Modern Languages and Literature, 48(3), 25–39. https://doi.org/10.17951/lsmll.2024.48.3.25-39
Anikó Németh
Budapest University Of Technology And Economics
Die Akteure der globalisierten Welt kommunizieren bereits seit geraumer Zeit in einer gemeinsamen, auf einer Vereinbarung basierenden Lingua franca. In der heutigen Zeit wird diese Funktion von der englischen Sprache übernommen, die als "English as a lingua franca", kurz "ELF", bezeichnet wird (Nickerson 2005, Ehrenreich 2010). Gemäß den Forschungsergebnissen (Pitzl 2010) wird der Ausgang von Geschäftsverhandlungen in hohem Maße von der Wahl der Sprache sowie der Kompetenz der Verhandlungsparteien, mit dem Verhandlungsinstrument Sprache umzugehen, beeinflusst.
Geschäftliche Verhandlungen und Unternehmensbesprechungen mit Dolmetschern stellen jedoch bislang ein unerschlossenes Forschungsgebiet der Sprachwissenschaft dar. In der Fachliteratur zur Übersetzungswissenschaft existieren nur wenige Studien, die sich mit dem Anforderungsprofil für Dolmetscher/innen in der Geschäftswelt sowie mit der Erfassung der Kundenbedürfnisse und -erwartungen an Verhandlungsdolmetscher/innen befassen (Karanasiou 2017, Zheng-Xiang 2018).
In meiner Präsentation werden die Ergebnisse einer qualitativen empirischen Studie präsentiert, die die Bedürfnisse und Erwartungen von Führungskräften internationaler Großunternehmen im Zusammenhang mit Verhandlungsdolmetschen beleuchten. Die Ergebnisse der Studie legen nahe, dass die Dienste von Sprachmittlern nach wie vor benötigt werden, wenngleich sich die Art der Aufgaben oder die Zusammensetzung der Sprachen von Zeit zu Zeit ändern. Neben der sprachlichen Kompetenz der Dolmetscher/innen sind das Vertrauen sowie zwischenmenschliche Beziehungen von herausragender Bedeutung beim Dolmetschen im geschäftlichen und unternehmerischen Kontext.
In meinem Vortrag werde ich auf die Einschätzung der gegenwärtigen Entscheidungsträger in der Wirtschaft zur zukünftigen Rolle von menschlichen Dolmetschern bei Geschäftsverhandlungen eingehen. Darüber hinaus wird erörtert, ob diese besonders sensible Dienstleistung als Mehrwert oder als notwendiges Übel betrachtet wird. Die Ermittlung der Kundenbedürfnisse ist einerseits für den Beruf des Sprachmittlers und andererseits auch für die Dolmetscher/in Ausbildung von herausragender Bedeutung.
References:
Ehrenreich, S. (2010). English as a business lingua franca in a German multinational corporation: Meeting the challenge. The Journal of Business Communication, 47(4), 408–431. https://doi.org/10.1177/0021943610377303
Karanasiou, P. P. (2017). Fulfilling the interpreting mandate in business negotiation meetings: The perspectives of interpreters and clients (Unpublished doctoral dissertation). Heriot-Watt University, Edinburgh, UK.
Nickerson, C. (2005). English as a lingua franca in international business contexts. English for Specific Purposes, 24(4), 367–380. https://doi.org/10.1016/j.esp.2005.02.001
Pitzl, M.-L. (2010). English as a lingua franca in international business: Resolving miscommunication and reaching shared understanding. VDM Verlag Dr. Müller.
Zheng, B., & Xiang, X. (2018). Between invisibility and over-visibility: Self-perception and user expectations of liaison interpreters in business settings. Babel, 64(1), 1–32. https://doi.org/10.1075/babel.00023.zhe
Chantal Wright
School Of Applied Linguistics ZHAW
Der grosse Wandel durch NMT liess die Welt der Literaturübersetzung weitgehend unberührt. Erst seit der Ankunft von GenAI im Jahr 2022 stellen sich Literaturübersetzer:innen die Frage, welche Konsequenzen diese Entwicklung für ihren Beruf hat. Die ersten Verlagshäuser und Agenturen (Taylor & Francis, Veen Bosch & Keuning, GlobeScribe) kündigen bereits Übersetzungen mit KI an. In der Forschung innerhalb der Übersetzungswissenschaften konzentriert man sich auf die Übersetzbarkeit literarischer und komplexerer Texte mit NMT, ohne sich mit philosophischen Fragen auseinanderzusetzen. Innerhalb des Berufs stossen die neuen Entwicklungen auf Ablehnung und Angst. Wir brauchen dringend sowohl eine klare Philosophie unserer literarischen Übersetzungspraxis als auch die Bereitschaft, die Möglichkeiten der KI zu untersuchen, wie es die deutsche Kollektive Kollektive Intelligenz tut, und womöglich und wo nützlich in unsere übersetzerische Praxis zu integrieren. In diesem Beitrag werde ich argumentieren, dass wir das literarische Übersetzen nicht durch KI ersetzen lassen sollen, dafür ist es eine zu wesentliche und ja „embodied‟ menschliche Tätigkeit —, wir müssen aber den neuesten technologischen Entwicklungen gegenüber offen sein. Was wäre, wenn GenAI genau der Gesprächspartner wäre, den wir Übersetzende für unsere oft einsame Tätigkeit brauchen?
Agnieszka Stawikowska-Marcinkowska und Jacek Makowski
University of Lodz
Der Beitrag präsentiert das Konzept sowie erste Zwischenergebnisse einer fachsprachenorientierten, textlinguistisch und medienlinguistisch ausgerichteten vergleichenden Studie zum Phänomen der sprachlichen Vielfalt von Kochrezepten im audiovisuellen Genre der kulinarischen Reisen. Der Schwerpunkt liegt dabei auf traditionellen und regionsspezifischen Kochrezepttexten sowie deren audiovisuellen multimodalen Realisierungsformen. Die charakteristischen textexternen und textinternen Variationsparameter der Textsorte Kochrezept werden im vorliegenden Beitrag an gewählten Sprachbeispielen aus den YouTube-Kanälen von @Robert_Maklowicz, @Książulo und @Lukasgalgenmueller kontrastiv dargestellt und erläutert.
Marcin Michoń
University of Lodz
Fokus auf die Wahrnehmung der Bedürfnisse der Adressat*innen der übersetzten Texte ist eine Kompetenz, die weiterhin eher dem Menschen als der Künstlichen Intelligenz eigen ist. Die Umsetzung der Skopos-Theorie bedeutet im Hinblick auf mannigfaltige Kommunikationsformen besondere Konzentration auf die Immersion. Diese (bewusste oder unbewusste) Reaktion der Empfänger*innen steht Mittelpunkt des Beitrags und das Untersuchungsobjekt bildet ein Literarisches Werk samt dessen filmischer Adaptionen. Es wird die Frage danach beantwortet, inwieweit die Qualitätssicherung veröffentlichter Texte, die nach Jahren in der Gruppe ihrer Adressaten anderes als (ursprünglich) abgezielt, wahrgenommen werden, gesichert werden kann Die Studie konzentriert sich auf die kognitionsorientierten, konnotativen und semantischen Aspekten der Rezeption, sowie den immersiven Mehrwert eines literarischen Textes über 40 Jahre nach seiner Veröffentlichung. Dabei wird die Wahrnehmung des Textes beobachtet, analysiert und reflektiert. Wichte Impulse für die Erfassung der Schlussfolgerungen setzte im Gespräch der Autor der Übersetzung.
Łukasz Plęs
University of Lodz
Das Ziel des Vortrags ist es, das Problem der zweifachen Fachlichkeit (S. Göpferich) vor dem Hintergrund der Fachübersetzung, insbesondere der Patentübersetzung, zu beleuchten. Einerseits zeichnen sich Patentschriften durch genaue und nachvollziehbare Beschreibungen aus, andererseits müssen sie dermaßen diffus verfasst werden, dass die Konkurrenz daraus keine eigene Entwicklung direkt ableiten kann. Zu ihren typischen Merkmalen gehört eine Vielfalt stilistischer Gepflogenheiten sowie terminologischer Besonderheiten. Die zweifache Fachlichkeit der Patentschriften ist darauf zurückzuführen, dass sie zum einen technische Sachverhalte betreffen und zum anderen diese Sachverhalte in einer juristischen Funktion darstellen. Aus diesem Grund sind Patentschriften in vieler Hinsicht eine Herausforderung für Übersetzer, da Nichteinhaltung patentrechtlicher Konventionen folgenschwer sein, d.h. den Schutzbereich eines Patents beeinträchtigen kann.
Krzysztof Sakowski
University of Lodz
Barcelonas Aufsatz (2017), in dem er davon ausgeht, dass vollständige Wörter auf dem Wege einer Metonymiesierung zu Kurzformen reduziert werden, indem nur saliente (d.h. auffällige, erkennbare) Elemente zurückgelassen werden, wird hier als methodologische Grundlage für eine kontrastive Analyse zwischen Polnisch und Deutsch betrachtet, die mögliche Unterschiede in der Prägnanz gewisser Elemente ans Licht bringen sollte. Dabei werden von Barcelona gewisse Faktoren der Salienz vorgeführt, deren Kombination zu bestimmten Wortsegmentierungsprinzipien führt, die im Weiteren an einem von ihm vorgeschlagenen Korpus von medizinischen Kurzwörtern analysiert werden. Ich möchte seine These an einem anderen, zweisprachigen Korpus überprüfen. Literatur: Barcelona, Antonio (2017): Salience factors determining natural metonymic clippings illustrated through the medical lexicon. In: Ibérica 34 (2017): S.17-44.