Blog CRC1646

Nachlese: Unsere Wissenshappen zu „Autor: ChatGPT?“ in der Wissenswerkstadt Bielefeld

2026-04-14T18:58:29+02:00

Bei dieser Veranstaltung, die am 8. November 2025 im Rahmen der Geniale „KI und wir“ stattfand, haben wir gezeigt, wie im SFB zu menschlicher und künstlicher Kreativität geforscht wird. Es gab drei Stationen, in denen unsere Forschenden mit den Besucher*innen über folgende Themen ins Gespräch kamen:

Wie unterschiedlich ist das sprachliche Verhalten von Menschen und ChatGPT und Co.? In der ersten Station haben Florian Kankowski & Prof. Dr. Oliver Bott (beide Projekt Ö) erläutert, wie im SFB mit einfachen Textfortsetzungsaufgaben untersucht wird, wie KI Sprache versteht, verarbeitet und selbst erzeugt.
Wie messen wir sprachliche Kreativität in literarischen Werken? In der zweiten Station diskutierten Emilie Sitter & Prof. Dr. Berenike Herrmann (beide Projekt A05) an Beispielen mit dem Publikum, was kreative Sprache in der Literatur überhaupt ausmacht.
Wie erzeugt KI aus Texten gesprochene Sprache? In der dritten Station verglichen Dr. Birte Schaller (Projekt Ö) & Prof. Dr. Hendrik Buschmeier (Projekt B02) gemeinsam mit den Teilnehmer*innen menschliche mit KI erzeugten Podcasts und verständigten sich darüber, ob bzw. wie diese Technik sinnvoll eingesetzt werden kann.

Für alle Interessenten haben wir zwei aktuelle Angebote, ein digitales und ein analoges:

Die Sprachschmiede: unser Citizen Science Sprachlabor rund um sprachliche Kreativität zum Mitforschen.
Brainstorm in der Wissenswerkstadt „Zwischen den Zeilen. Ein Abend über die verborgenen Strukturen unserer Sprache“ mit Ralf Vogel & Ina Pick am 22. April, 19.30 Uhr, Tickets gibt es über die Wissenswerkstadt

Graphic Recording der Veranstaltung von Moderatrix, © SFB 1646

Die „Sprachschmiede“. Das Citizen Science Sprachlabor des SFB 1646 ist online

2026-04-08T11:24:00+02:00

Die "Sprachschmiede" ist eine neue Web-Plattform zu linguistischen Phänomenen rund um sprachliche Kreativität in der Kommunikation. Sie wurde von Florian Kankowski, Doktorand im Projekt Ö, mitentwickelt. Die Projektleitung hat Prof. Dr. Oliver Bott. Sie soll es interessierten Bürger*innen ermöglichen, sich auf Augenhöhe an der Forschung im SFB zu beteiligen. Dabei ergänzt sie Projekte der partizipativen Sprachforschung, die der Praxis des Linguistic Landscaping folgen, indem Citizen Science Ansätze nun auch in der linguistischen Grundlagenforschung zum Einsatz kommen.

Das Besondere an der „Sprachschmiede“ ist ihr inhärent partizipatorischer Ansatz. Die Citizen Scientists können ihren Interessen nachgehen und an verschiedenen Stellen des wissenschaftlichen Prozesses mitwirken: von der Datenakquise durch Teilnahme an spielerisch gestalteten Experimenten über komplexere Analyseaufgaben bis hin zur Formulierung eigener Hypothesen, die sie gemeinsam mit den Forschenden in Forenbeiträgen diskutieren können. Im „Sprachblog“, dem hauseigenen Blog der Plattform, werden Ergebnisse aus Forschungsprojekten diskutiert und für die Öffentlichkeit zugänglich gemacht.

Die Sprachschmiede macht sich zunutze, dass uns das Thema Sprache ständig umtreibt. Unabhängig davon, ob wir uns bewusst damit beschäftigen oder nicht, haben wir einen intuitiven Zugang dazu. Mit Hilfe der Plattform erfahren die Forschenden von implizitem sprachlichen Wissen, das sich auf diese Weise viel umfassender erforschen lässt als in einer klassischen Laborsituation, auf die psycholinguistische Studien üblicherweise ihre Daten stützen. Außerdem bietet sie viele Möglichkeiten, um für sprachliche Phänomene im Alltag zu sensibilisieren und Einblicke in wissenschaftliche Fragen und Methoden zu geben. Dabei wird das Mittel der Gamifizierung für die Gestaltung von Experimenten genutzt, ein Ansatz, der in der Wissenskommunikation zunehmend Anwendung findet. So können sich die Teilnehmenden auf der Plattform für verschiedene Spiele entscheiden.

Aktuell laufen zwei Citizen-Science-Projekte auf der Sprachschmiede: In „Mensch vs. Maschine“ begeben sich die Citizen Scientists auf eine Forschungsreise zum Thema KI-generierter Sprache. Dabei lernen sie in mehreren Quizzen, diese von menschlicher Sprache zu unterscheiden. Im digitalen Escape-Room „Verschollen im Nichts“ befassen sie sich dagegen mit sprachlicher Negationsverarbeitung. Hier werden aus Sprachexperimenten Sprachspiele und Rätsel, ohne ihren wissenschaftlichen Mehrwert einzubüßen.

Die Plattform wird laufend weiterentwickelt und mit neuen „Forschungsreisen“ rund um sprachliche Kreativität ergänzt.

Siehe hier auch die Pressemitteilung der Universität Bielefeld vom 8. April 2026.

Launch der "Sprachschmiede" im SFB Kolloquium, Foto: Sascha Hermannski © SFB 1646

Screenshot einer Aufgabe in der "Sprachschmiede"

oben links: Florian Kankowski, oben rechts: Oliver Bott.

unten: Team des Projekts ö

Fotos: Sascha Hermannski © SFB 1646

Konferenz-Rückblick: Projekt A05 in der DHd2026 in Wien

2026-03-30T14:45:53+02:00

Die DHd ist die größte Konferenz in den Digital Humanities im deutschsprachigen Raum, wo in der Forschende aus den verschiedensten Disziplinen der Geisteswissenschaften zusammenkommen, die mit digitalen Methoden und Tools arbeiten. Entsprechend anregend und vielseitig gestalten sich dort Austausch und Perspektiven. Während die DHd2025 in Bielefeld ausgerichtet wurde, reisten nun Emilie Sitter, Karoline Rehage und Berenike Hermann vom Projekt A05 zur DHd2026 nach Österreich. Die Konferenz fand vom 23. bis 27. Februar 2026 im besonderen Ambiente der Universität Wien statt und wurde u. a. von Prof. Dr. Tara Andrews und Prof. Dr. Thomas Wallnig organisiert, bei denen Emilie während ihres Masters in Wien studiert hat.

Auf zwei intensive Workshop-Tage, an denen neue, praktische Skills erworben werden konnten, folgten drei Konferenztage mit zwei Keynotes, parallelen Vorträgen, Podiumsdiskussionen und Postersessions. Durch das breit gefächerte Programm konnten sich die Projektmitglieder von A05 auf den neuesten Forschungsstand in den Digital Humanities bringen. Veranstaltungen, die den Computational Literary Studies zuzuordnen sind, waren deshalb besonders relevant.

Das diesjährige Motto der Tagung lautete „Nicht nur Text, nicht nur Daten“. Da A05 fast ausschließlich mit Text und Daten arbeitet, nahm das Projektteam das Motto für ein Poster mit dem Titel „Mehr Text! Mehr Daten!“ zum Anlass. Im eleganten Festsaal der Universität Wien stellten Emilie und Karoline das „Kreativitätsduell“-Spiel vor. Es wurde in Zusammenarbeit mit dem Projekt Ö für die Bielefelder Geniale 2025 entwickelt, um die Besucher*innen der Bielefelder Wissenswerkstadt auf spielerische Art zur Auseinandersetzung mit sprachlicher Kreativität anzuregen, und kann noch heute dort als Mitmach-Station genutzt werden.

In der DHd2026 konnten die Konferenzteilnehmer*innen vor Ort das Spiel ausprobieren und selbst herausfinden, ob sie „kreativer“ als ChatGPT texten. Poster und Demo führten zu vielen spannenden Diskussionen rund um Gamification, aber auch zu grundsätzlichen Reflexionen über sprachliche Kreativität.

Ein besonderes Highlight für A05 war die Eröffnungskeynote von Miriah Meyer (University Linköping) zur Pluralität von Daten, da sie sich in ihrem Vortrag der spielerischen Vermittlung von Erkenntnissen aus Daten für Menschen jeden Alters widmete: von deren Analyse und Visualisierung bis zur grundlegenden Frage, was Daten überhaupt sind. Ihre kreativen Ansätze zu diesem Thema waren anschließend in aller Munde. Für A05 war ihr Beitrag besonders interessant, weil das „Kreativitätsduell“ auch aus der Idee heraus entstanden ist, niederschwellig und spielerisch Daten für die Kreativitätsforschung zu generieren.

Die Teilnahme an der DHd2026 war deshalb in mehrfacher Hinsicht eine wertvolle Erfahrung. Zum einen handelte es sich für Karoline, die als wissenschaftliche Hilfskraft im Projekt tätig ist, um ihre erste Konferenz, auf der sie an einem Beitrag beteiligt war. Zum anderen bekundeten viele Besucher*innen der DHd ein großes Interesse am Thema Kreativität, auch wenn sie selbst (noch) nicht dazu forschen. Schließlich knüpfte das Projekt A05 neue Kontakte, aus denen Kooperationen entstehen können, so etwa im Hinblick auf spannende Daten, die sich auf die im Projekt entwickelten Metriken zur Messung sprachlicher Kreativität anwenden lassen. Fest vereinbart ist bereits ein Treffen mit dem Team des an der Universität Darmstadt angesiedelten Liebesbriefarchivs, um sich über kreative Kosenamen in Liebesbriefen auszutauschen. Es begann 2021 unter dem Titel „Gruß & Kuss“ selbst als ein sprachwissenschaftliches Citizen Science Projekt.

Gruppenfoto der AG Herrmann

v. l.n.r. Emilie Sitter (PhD in A05), James Wiebe (wissenschaftliche Hilfskraft in der AG Herrmann), Tobias Duda (studentische Hilfskraft in der AG Herrmann), Berenike Herrmann (PI von A05), Karoline Rehage (wissenschaftliche Hilfskraft in A05), Pierre Michel Weiße (wissenschaftliche Hilfskraft in der AG Herrmann), Marie-Christine Boucher (Postdoc in der AG Herrmann), Robin-Martin Aust (Postdoc in der AG Herrmann)

Foto: Frank Fischer © FU Berlin

Emilie Sitter und Karoline Rehage vor ihrem Poster

Foto: Pierre-Michel Weiße © Uni Bielefeld

Projektteam A05

v.l.n.r.: Sina Zarrieß, Omar Momen, Emilie Sitter, Berenike Herrmann

Foto: Sascha Hermannski © SFB 1646

Conference Review: Amelie Ewald (C01) at the CELISO’s annual doctoral colloquium in Paris

2026-03-18T12:26:19+01:00

On February 6, Amelie Ewald (project C01) contributed to the CELISO’s annual doctoral colloquium by presenting her dissertation project, along with eight other doctoral candidates. CELISO – The Sorbonne Centre for Linguistics – brings together the university’s researchers specialising in the Germanic, English, Scandinavian, and Slavic languages. The colloquium was organized by Brahim Bougnouch, Maëline Buge, Martine Etchart and Mélanie Gantier and gave PhD students affiliated to various French and international universities the opportunity to present their work and progress. In addition to PhD students, lecturers from CELISO were in attendance and engaged in lively discussions. Presentation sessions alternated with breaks, during which participants continued to discuss their topics, talked about their life as a PhD student outside their home universities and/or countries, encouraged one another and shared tips.

In her talk, Amelie Ewald explored ideas for experimental studies for investigating cognitive, linguistic and biographic foundations of wordplay affinity and examined the relevance for linguistic creativity. She was delighted by the positive feedback and greatly appreciated the warm, friendly exchange. An invitation to visit Bielefeld as part of the 11^th International Conference of the German Cognitive Linguistics Association has already been accepted; perhaps some more participants will be drawn to Bielefeld even earlier for the DGfS Summer School.

Link to the program of the CELISO’s annual doctoral colloquium 2026.

Paris, photo: Amelie Ewald © SFB 1646 Sorbonne, photo: Amelie Ewald © SFB 1646

Amelie Ewald, photo: Johannes Ewald © Amelie Ewald

Project C01, photo: Sascha Hermannski © SFB 1646

back row from left to right: Joana Cholin, Lisa Hudalla & Amelie Ewald

front row from left to right: Jana Häussler & Dilara Balaban

Conference Review: Julia Demina (B03) at the workshop „Social Meaning and Grammar“ in Zurich

2026-03-17T16:40:36+01:00

The workshop „Social Meaning and Grammar“ was organized by Laura Reimer and Andreas Trotzke and took place on 10-11 February 2026 at the University of Zurich. It invited contributions from formal linguistics frameworks analysing social meaning components, sociolinguistic frameworks focusing on linguistic form, and experimental work on processing social meaning, including questions regarding acquisition, comprehension, and production. This interdisciplinary focus made the workshop an ideal venue for project B03 to present its ongoing research on indirectness in discourse.

The workshop brought together researchers working at the intersection of syntax, semantics, pragmatics, and sociolinguistics. The programme featured keynote presentations and poster sessions exploring how grammatical structures encode social meaning. Topics ranged from formal analyses of politeness markers to experimental investigations of how speakers process socially loaded linguistic forms. The workshop provided a valuable opportunity to engage with cutting-edge research on the grammatical encoding of social information across different languages.

Julia Demina presented the research of B03 in a poster entitled „Indirect Requests as Grammaticized Social Updates: A Cross-linguistic Study“. The poster presented the theoretical framework and methodology of the project, which investigates how indirect requests function as performative social updates that are syntactically encoded in grammar. Drawing on Krifka's (2023, 2024) commitment-based syntax and Wiltschko's (2016, 2021) Interactional Spine Hypothesis, the research explores how interactional moves can result in social updates when they are conventionalized and socially binding.

The poster outlined the project's cross-linguistic methodology based on Discourse Completion Tasks (DCT) in written form, following the CCSARP tradition (Blum-Kulka et al. 1989). The study examines three typologically diverse languages: Russian, Chinese, and Turkish, with parallel examples demonstrating ability-based indirect requests in each language. The research design manipulates two social variables: relative power relations between speaker and hearer, and the level of imposition involved in the request. With data from 145 participants per language, the project tests whether interactional context triggers specific grammatical forms, thereby providing empirical evidence for the hypothesis that social meaning manifests as stable correspondences between interactional context and grammatical structure.

A key innovation presented was the project's three-layer annotation scheme: morphosyntactic annotation, pragmatic strategies, and modification patterns. The main principle of morphosyntactic annotation follows a TYPE-FORM-REALIZATION hierarchy, enabling systematic cross-linguistic comparison of how different languages grammatically encode indirect speech acts.

During the poster session, colleagues provided valuable feedback on the theoretical integration of commitment-based approaches with the Interactional Spine Hypothesis. Discussions with other researchers working on speech act syntax and social meaning yielded constructive suggestions for extending the analysis and potentially led to future collaborations. The cross-linguistic dimension of the project, comparing structurally and typologically diverse languages, was particularly appreciated by participants interested in the universality and language-specificity of social grammar.

Overall, participation in the „Social Meaning and Grammar“ workshop allowed project B03 to introduce the CRC 1646 to the social grammar research community, to foster relationships with researchers working on the syntax-pragmatics interface, and to deepen our understanding of how grammar encodes social meaning. The workshop's focus on innovative approaches to social grammar directly aligns with the CRC's core goals of investigating linguistic creativity in communication and the mechanisms by which speakers create and process indirect speech acts.

Zurich, photo: Julia Demina © SFB 1646

Julia Demina (B03), photo: Sascha Hermannski, © SFB 1646

Project B03, photo: Sascha Hermannski, © SFB 1646

back row from left to right: Raoul Schubert & René Nicolas

front row from left to right: Arndt Riester, Jutta Hartmann, Tanja Ackermann & Julia Demina

A Short Report of Research Stay at KU Leuven by Désirée Kleineberg (A06)

2026-03-06T07:41:48+01:00

From 12^th of November to 9^th of December 2025, the postdoctoral fellow Désirée Kleineberg from project A06 profited from the outgoing scholarships of CRC 1646 and seized the opportunity for a research stay at the KU Leuven (Belgium). She was hosted by the research group of Functional and Cognitive Linguistics (FunC). This research group investigates grammar from a cross-linguistic perspective, focusing on form-function relations, the general architecture of grammatical systems and diachronic evolution of grammatical constructions. With this research focus, the FunC group was a perfect match for Désirée’s research on the evolution of verbal periphrases in the Romance languages. In her current studies, she concentrates on the evolution of competition between various constructions which have the same function, and possible outcomes of those processes – in short: the question why some linguistic innovations are successful and others not.

During her stay, she had the opportunity to discuss her studies, the theoretical modelling and methodological questions with colleagues like Hendrik De Smet, Bert Cornillie, Anna Kisiel, or María Sol Sansiñena. The fact that we were able to welcome Hendrik De Smet to the Bielefeld colloquium just one week after Désriée's stay to give a lecture on "In language change, analogy is like gravity: omnipresent, weak and poorly understand" is proof that the connections between the CRC 1646 and our colleagues at the KU Leuven are manifold and exist in both directions. On the 18^th of December she presented the CRC and her work in the FunC-Colloquium (“On “failed” changes – The (rise and) decline of Fr. ”) and on the 3rd of December in a student’s course on language variation and change in French (“Évolution de la périphrase - Analyse des facteurs contextuels de son déclin du 15e au 18e siècle)”.

Luckily, there was also time to explore the beautiful historical city of Leuven, make some trips to nearby cities and to eat one or the other Belgian waffle!

After a month which was packed full with fruitful discussions, productive writing and some thorough reflections on theoretical concepts like competition, differentiation and success of linguistic creativity, Désirée came back full of inspiration and motivation to continue her work in project A06.

Central University Library, Foto: Désirée Kleineberg © SFB 1646

River Dijle at Groot Begijnhof, Foto: Désirée Kleineberg © SFB 1646

Désirée Kleineberg, Foto: Sascha Hermannski © SFB 1646

Project A06, Foto: Sascha Hermannski © SFB 1646

Hintere Reihe, v.l.n.r.: Barbara Job, Tanja Ackermann, Said Sahel

Vordere Reihe, v.l.n.r.: Rebekka von Legat, Désirée Kleineberg

Was macht der SFB 1646 eigentlich? Start unseres YouTube Kanals

2026-03-02T16:34:15+01:00

Womit sich unser Sonderforschungsbereich befasst, erfährt man an seinem Titel, nämlich mit: „Sprachlicher Kreativität in der Kommunikation“. Aber: Was heißt das konkret?

Stellvertretend für das Sprecherteam des SFB berichtet im ersten Video in unserem YouTube Kanal Ralf Vogel, Professor für germanistische Linguistik, davon, was uns an sprachlicher Kreativität interessiert, mit welchen Methoden wir sie erforschen und warum dabei der Austausch über die alltägliche Sprachpraxis besonders wichtig ist.

Das vollständige Video findet sich hier.

Demnächst erscheinen dort auch die ersten Teile unserer Serie „Behind the Science“ mit verschiedenen Mitarbeitenden des SFB sowie Mitschnitte aus unseren öffentlichen Veranstaltungen u.a. in der Wissenswerkstadt und der VHS Bielefeld.

Infos zu Ralf Vogel.

Mehr zu den Projekten des SFB hier.

Still aus dem Video © Sascha Hermannski, SFB 1646

Conference Review: The CRC 1646 hosted the SemDial 2025 „Bialogue“ in Bielefeld

2026-01-15T21:40:06+01:00

SemDial returned to Bielefeld after 24 years. The CRC 1646, together with the TRR 318 “Constructing Explainability” and the Center for Cognitive Interaction Technology (CITEC), hosted the 29th Workshop on the Semantics and Pragmatics of Dialogue, which took place from 3rd to 5th September 2025. Under the theme "Meaningful Interaction", SemDial 2025 had a special focus on empirical, theoretical, and computational work that investigates the prerequisites for meaningful interaction, such as cognitive mechanisms, computational representations, but also human users’ attitudes or ethical considerations.

The PIs Hendrik Buschmeier (B02, INF), Stefan Kopp (C02), Sina Zarrieß (A05, B02) und the Postdoctoral Fellow Maryam Mohammadi (INF) were part of the local organising committee. The three-day workshop consisted of three keynotes, five sessions with presentations and two with posters. The keynotes were given by Arabella Sinclair (University of Aberdeen) on “The Many Reasons for Repetition in Dialogue”, Robert D. Hawkins (Stanford University) on “Foraging for Common Ground” and – invited by the CRC – David Schlangen (Potsdam University und CRC 1287) on “Meaningful Interaction with Unreal Speakers”.

Contributions from the CRC:

Leonie Schade presented experimental work from the A02 project on the production of novel phonetic forms and the innovative methodology to assess them. In her talk “Does it take two to do an articulatory tango? Investigating the production of novel phonetic forms in varying communicative settings” co-authored with Daniel Duran, Sina Zarrieß, Joana Cholin and Petra Wagner, she discussed how speech dynamics depend on the communicative setting, social context and individual differences in speakers’ personality.

Bastian Bunzeck from the A02 project presented a poster entitled “Child-Directed Speech Is Fine-Tuned to Children’s Developmental Needs” based on earlier work in collaboration with Holger Diessel (FSU Jena). Crucially, their work highlights the importance of the communicative setting and how it affects variation in situated interaction. Their data neatly “show that the relative frequency of the different CDS constructions varies with the social situation and children’s age”.

Shravani Patil from the project A03 presented two posters on the particle ‘-hii’ in Hindi based on her joint work with Farhat Jabeen. The first was entitled “One -hii to rule them all: Functions and intonation of particle -hii in Hindi movie dialogues“. Here they present a movie corpus for the analysis of ‘-hii‘ and propose that this particle, despite having several context-dependent interpretations, has a single lexical entry: contrast. The second was entitled „Intonation and prosodic phrasing of particle ‘-hii’ in Hindi/Urdu dialogues” and showed the first evidence of variability in the attachment of ‘-hii’ in a nominal phrase. René Nicolas and Arndt Riester (B03) presented their work on response strategies of politicians in interview settings with a special interest in evasive strategies. In her joint Poster entitled “Towards a discourse-structure based response classification for interviews” they proposed a response classification and showed some examples of the different categories taken from a corpus of QUD-annotated interviews.

Finally, Sina Zarrieß (A05, B02) contributed to a Poster by her research fellows Judith Sieker and Clara Lachenmaier entitled “The FLEX benchmark: Evaluating pragmatic competence in LLMs via false presuppositions”, in which they highlight critical limitations in current LL.Ms’ ability, especially regarding grounding behaviour and the influence of context and political orientation.

For the CRC 1646 Participants of the SemDial 2025 “Bialogue” was a great success. The stimulating discussions on topics relating to the communication of meaning in an interactive dialogue between humans and machines helped to advance the projects.

Leonie Schade (A02) on the right side Bastian Bunzeck (A02)

on the right side Shravani Patil (A03) René Nicolas (B03)

all © Sascha Hermannski, SFB 1646

Conference Review: Project A03 organised Workshop on “Mismatches in Information Structure, Prosodoy, and Syntax: Challenges and Chances in the Data”

2026-01-05T12:16:49+01:00

The workshop “Mismatches in Information Structure, Prosody, and Syntax: Challenges and Chances in the Data”, organized by the members of project A03 at Center for Cognitive Interaction Technology (CITEC) at Bielefeld University, focused on highlighting interesting and challenging mismatches in data based on information structure, syntax, prosody, and their interface. Ojaswee Bhalla (Indian Institute of Technology Delhi), Frank Kügler, Corinna Langer (both Goethe-Universität Frankfurt a. M.), Tatiana Luchkina (Stony Brook University), Katalin Mády, Balázs Surányi (both Research Institute for Linguistics Budapest) were among the several specialized researches in the field who were invited speakers for the workshop. The programme brought together a diverse set of languages, including Hindi/Urdu, Hungarian, German, Russian, and English.

The event departed from the usual workshop format. Instead of full-length talks, each contribution consisted of a concise 5–10 minute lightning presentation centered on a small set of data, followed by an extended 35–40 minute discussion. The two day workshop opened with Jutta Hartmann and Petra Wagner (both from A03), who provided an overview of the goals of project within the broader context of the CRC’s theme of linguistic creativity.

The first day concentrated on syntactic perspectives on information-structural phenomena, while the second day turned to prosodic and phonetic approaches. Shravani Patil (A03) presented an analysis of the Hindi/Urdu particle -hii based on a Bollywood film corpus, which was followed by Ojaswee Bhalla’s contribution on the discourse particle -to. These talks paved a way for interesting discussions not only based on information structure and syntax, but also on the notorious nature of discourse particles in Hindi/Urdu. Tatiana Luchkina presented on integrating prosodic and structural cues in Russian and Hindi/Urdu. The first day concluded with Balas Suranyi’s talk on marking broad focus.

The second day of the workshop focused on prosodic and phonetic aspects of information structure. Contributions by Katalin Mády, Ákos Buza, Farhat Jabeen, Corinna Langer, and Frank Kügler addressed variability in prosodic realization, experimental approaches to prosody–information structure mismatches, and their implications for theoretical analysis. The discussion highlighted the challenges of defining stable baselines and interpreting prosodic variation across speakers and structures.

The workshop facilitated closer methodological and theoretical exchange with invited speakers, especially on issues of experimental design and cross-linguistic comparison. At the same time, discussions of both cross-linguistic and within-language variation contributed to a clearer theoretical perspective and baseline for the project.

Workshop programme

Project A03

Project A03 © Sascha Hermannski, SFB 1646

Conference-Review: Project A01 organised Workshop on “Cross-clausal Dependencies”

2025-12-19T16:25:15+01:00

The workshop “Cross-clausal Dependencies (CCD)” aimed to shed light on questions regarding the syntactic configuration of clauses involving long-distance dependencies, the properties of domains involved in long- distance dependencies, as well as the role of pragmatic and semantic context for the acceptability of long-distance dependencies. During the three-day workshop, 14 presentations addressed these questions and provided insights based on different phenomena, theoretical backgrounds, and languages.

The workshop contained three keynote presentations. Krisztina Valeria Szécsényi (University of Deusto) discussed „Long-distance agreement in Basque and Hungarian“. Lutz Gunkel (IDS Mannheim) presented collaborative work on German prepositional object clauses together with Jutta M. Hartmann (see below). Renuka Ozarkar (University of Mumbai), a former visiting scholar of the CRC (see for her report), investigated interactions of „Clausality and long distance agreement in Marathi“ and Stefan Keine (UCLA) presented an analysis on long-distance dependencies as „Collateral movement“.

Prior to the workshop, members of A01 presented the project and current work. András Bárány and Jutta M. Hartmann introduced the project in their talk „The Role of Analogy in Generative Grammar“. In his talk „Reanalysis by Analogy: Extraction from German wenn-clauses“, Fabian Zöfelt discussed experimental data and an analysis on long-distance dependencies into complement- like wenn ‘when’-clauses in German. Szilvia Daczó presented (new) data on the controversial topic of long-distance agreement with intransitives in Hungarian. Her talk was entitled „Annual review on long-distance object-verb agreement with intransitive matrix verbs in Hungarian“. Jutta M. Hartmann and Lutz Gunkel discussed potential reanalyses of prepositional object clauses without an overtly realized preposition in their contribution „Extraction from prepositional object clauses: cross-linguistic, empirical and theoretical insights“. Anke Himmelreich spoke about a theoretical puzzle in Mabia languages in her talk „When long-distance dependencies are actually short: The case of Mabia languages“.

Submitted papers were presented by Klaus Abels (University College London), Dawhah Alruayni (University of York), Daniel Aremu (Goethe University, Frankfurt), Shin-Sook Kim (University of York) & Gereon Müller (Leipzig University), Irina Stoica (University of Bucharest), and Qiuhao Charles Yan (Queen Mary University of London).

The workshop was very helpful for us because it gave us a broader overview of current empirical and theoretical research dealing with the phenomena of long-distance dependencies and cross-clausal relations, which are relevant to our work. The CCD also helped us advance our own research, above all by providing us with helpful comments and insights from colleagues who are working on similar phenomena or have expertise in the theory. We would like to mention in particular our invited speakers, Krisztina Szécsényi and Stefan Keine, but also Gereon Müller, who recently referred to our work in his lecture at the Linguistics Colloquium in Bielefeld.

Conference Website & Program

Project A01

Conference Review: Florian Kankowski (Ö) at the CogSci 2025 in San Francisco

2025-12-12T15:06:41+01:00

Florian Kankowski had the opportunity to present his research at the CogSci 2025 conference in San Francisco. Organized by The Cognitive Science Society, the conference, titled "Theories of the past, theories of the future," took place from July 30th to August 2nd. The conference was structured with a diverse program of talks, symposia, and poster sessions. As is tradition for CogSci, the topics language, communication and cooperation were strongly represented at the conference, majorly intersecting with the research topics of the CRC.

Florian Kankowski (Project Ö) presented two posters at the conference, both situated within the domain of discourse. The first, "Instruction tuning modulates discourse biases in language models" (Project Ö), co-authored with Torgrim Solstad, Sina Zarrieß and Oliver Bott, investigated how instruction tuning and model size affect the emergence of Implicit Causality biases in different language models. Their findings revealed that instruction tuning can induce biases that align more closely with human-like preferences, even exceeding them in some cases.

His second poster, "Linguistic Creativity affects Discourse Expectations related to Contiguity Relations but not Implicit Causality" (Projects B01 and Ö), was co-authored with Oliver Bott and Torgrim Solstad. This poster presented findings from an experiment on human participants, showing that while a "creativity on demand" task can lead to more elaborate discourse, it does not override strong lexical-semantic biases like Implicit Causality. This suggests that even in creative language use, humans tend to produce language that adheres to their preferred discourse structure and instead elect to express linguistic creativity through different means.

Attending a large, interdisciplinary conference like CogSci was incredibly valuable for Florian. It provided a unique opportunity to connect with researchers from different sub-fields of cognitive science and get a broader perspective on his work. The feedback he received on his posters was insightful and has already given him new ideas for future research. In line with the overarching questions of the CRC, he is particularly interested in extending the LLM research to develop a more robust benchmark for comparing human and machine creative language use.

alle © SFB 1646

Conference Review: Project C02 organised Workshop on “Multimodal Creativity in Speech and Gesture Production” at the ZIF

2025-12-05T15:29:11+01:00

The workshop “Multimodal creativity in speech and gesture production” addressed the question of how humans employ speech and gesture creatively in communicatively challenging situations, e.g., when conventionalized or commonly used communicative resources are insufficient. It featured presentations from projects C02, C03 and C04, as well as from invited guest speakers. It additionally included two targeted discussion sessions, one about operationalizing and measuring multimodal creativity, and one about cognitive and computational modeling of multimodal creativity.

The three talks by CRC projects all presented experimental results along with novel approaches to analyzing those results. A recurring theme throughout all three talks was the limited applicability of existing literature on creative gestures, and the challenge of extending the CRC’s working definition of linguistic creativity to the gestural channel. These issues were tackled directly in the discussion sessions, where the workshop participants split into groups to raise concrete problems, offer potential solutions, and share useful tools and techniques.

The talks by the invited guest speakers did not necessarily address linguistic creativity explicitly, but they were all nevertheless tremendously valuable and relevant to the workshop’s topics. James Trujillo (University of Amsterdam), Anna Palmann (University of Amsterdam) and Jiahao Yang (University of Bath) presented empirical investigations of multimodal communication in various communicatively challenging situations, e.g., in noisy environments, directly relating to the CRC’s research questions and hypotheses about the contextual conditions that support or require creative use of language. Angela Grimminger (Paderborn University) and Andy Lücking (Goethe University Frankfurt) presented empirical findings and formal representations, respectively, regarding relatively conventional multimodal productions, a necessary step to understanding deviation from such conventions, i.e., instances of genuine creativity.

Contribution from CRC-Members:

Lisa Gottschalk, Alon Fishman, Joana Cholin & Stefan Kopp (C02): Creativity in speech and gesture and its perception in humans and AI
Lotta Heidemann (C03): Coding creative gesture use by people with neurogenic language and communication disorders
Luyao Wang (C04): Multimodal strategies for resolving lexical challenges: The role of creative gestures in four- and five-year-old children Program & Abstracts

Program & Abstracts

Conference Review: Project B05 hosted a Research Workshop on Open Texture

2025-11-20T08:39:45+01:00

To deepen our understanding of phenomenon of open texture and broaden our theoretical horizons, project B05 hosted a two-day workshop on 18-19 September 2025 bringing together scholars working on open texture or, more broadly, semantic indeterminacy. Broadly, our aim was to collect fresh perspectives on the fundamental nature of open texture, and how it differs from related phenomena such as vagueness.

The workshop opened with a talk by Derek Ball (St. Andrews), who examined the implications of “Temporal externalism and open texture” — the idea that future facts can influence the present meaning of our terms — for the theoretical problem posed by open texture. His talk raised the possibility that open texture may, in fact, be epistemic rather than semantic in nature. These ideas sparked lively discussion in the subsequent Q&A section, and kept people talking well into the conference dinner.

Céline Henne (Amsterdam) took the podium with “Open texture and the function of conceptual analysis”, a talk rooted in the difficulties classical conceptual analysis faces when confronting the apparent semantic gaps of our concepts revealed in open texture cases. She proposed the notion of alien conceptual analysis, which deals with this open-texture porosity by considering the forms of life within which our concepts are embedded.

After a social lunch break, the second half of the Thursday programme drew increased interest from the wider CRC, with multiple members joining us for the afternoon talks, the first of which was Dan Lassiter’s (Edinburgh), titled “Open texture via social-meaning-first-metasemantics”. He argued that meanings can often be taken as pointers into networks of social norms, gaps within which are revealed in open texture cases. On this view, which invited much curious and critical input from the audience, open texture is, in fact, a problem of social coordination.

Following a coffee break, Daniel Gutzmann (Bochum) gave a talk titled “Mechanisms for semantic creativity: what we can learn from diachronic semantics”. He provided a detailed overview of the mechanisms behind diachronic semantic change (how meanings shift over time), the most theoretically intriguing part of which are the contextually salient relations establishing a bridge between source and target meanings. Such considerations will help us understand how open-texture predicates may (or may not!) shift meanings.

Friday’s programme began with a talk by our PI Christian Nimtz, titled “Open Texture. A view from vagueness”. Christian set out with the observation that both open texture and vagueness involve borderline cases, and argued that if we find what distinguishes open texture borderline cases from those rooted in vagueness, we may be able to get clearer on what being open textured may amount to. His suggestion for such a distinguishing feature: Borderline cases in vagueness (and the kindred phenomenon of ordinal gappyness) can be dissolved by variations of values along the predicate’s decisive dimensions of application. This is not true for OT borderline cases; they are stable under inner-dimensional variation, even when we are concerned with multi-dimensional predicates. The stability vanishes only if we consider not variation within dimensions, but of dimensions – which in turn yields a neat hypothesis of what distinguishes the properties of being vague from being open textured.

Justin D’Ambrosio (St. Andrews) followed with “Clouds and Credences: a probabilistic approach to underspecification”. His talk addressed felicitous underspecification, which occurs when speakers are under-specific in what they (literally) say, but communicate successfully regardless, such as when I find myself at an unenjoyable event and utter to my companion, “I’ve had enough”. Justin challenged classical common ground approaches to this issue, suggesting that listeners may operate with probabilistic credences rather than fixed contents in interpreting such utterances.

The workshop concluded with a talk by further project B05 member Mark Bowker, titled “Open texture with minimal propositions”. Mark too suggested an approach to open texture that lays it out as an epistemic phenomenon; as indecision arising from a clash of the sorts of processes we use to check whether a predicate applies (e.g. checking for manhood by beardedness). This might occur, for example, when we encounter a case for which several of the processes associated with the same predicate yield conflicting results, or when processes associated with predicates usually taken to be mutually exclusive each rule positively on the case.

Overall, the workshop was a stimulating and productive event. It offered reinforcement and challenge to the members of B05, sharpening existing lines of thought while revealing promising new approaches to the issue of open texture. The participations of members from the wider CRC led to intriguing interdisciplinary exchange, deepening the ties between our project-specific interests and the CRC’s broader research programme.

Conference participants © Sascha Hermannski, SFB 1646

Projekt B05 © Sascha Hermannski, SFB 1646

Conference Review: Project C01 hosted the International Workshop on “Psycholinguistics of Lexical Creativity”

2025-11-14T23:25:18+01:00

On 8-9 September 2025, Project C01 hosted a workshop on the cognitive processes underlying creative language use, titled “Psycholinguistics of Lexical Creativity”. The event brought together researchers from within the CRC1646 and invited experts from around the world to discuss how people produce and understand creative words and expressions, and how these abilities differ across speakers and contexts.

The talks covered a wide range of perspectives on linguistic creativity. The first keynote speaker, Prof. Dr. Alissa Melinger from University of Dundee talked about the ways speakers choose between different words that signify the same thing or are very similar in meaning. Prof. Dr. Rasha Abdel Rahman from Humboldt-University at Berlin then presented a talk on the role of multisensory, associative, contextual and socially driven meaning processing for creative communication. meaning processing for creative communication.

On the second day of the workshop, Dr. Solène Hameau, from Université catholique de Louvain, spoke about creative strategies in individuals with neurogenic language disorders, arguing in what way so-called “errors” in word production can provide valuable insights into the mechanisms of language and Dr. Britta Biedermann from Curtin University, talked about how different factors, such as multilingualism, can influence the word selection processes.

The PhD students and researchers working within the C-Area of the CRC (see below) complemented the keynote talks with the presentations on their ongoing work. The sessions were followed by lively discussions that encouraged the exchange of ideas and highlighted opportunities for future collaboration. The workshop offered an engaging platform for reflecting on how linguistic creativity emerges in different speaker groups and contexts. It provided valuable insights into the cognitive processes and resources that allow speakers to deviate from conventional language use, whether to be inventive, humorous, or to communicate effectively when concepts and words are not available or seem inadequate.

Presentations given by members of the CRC1646:

Alexandra Filimonova & Denise Gajda (C04): Communication Strategies at the Intersection of Clinical Linguistics and Second Language Acquisition

Alon Fishman & Lisa Gottschalk (C02): Non-codable Stimuli in Reference Games: Salient Features and How to Capture Them

Mareike Hartmann & Lotta Heidemann (C03): Creativity in Verbal and Multimodal Communication of People with Neurogenic Language and Communication Disorders

Valentina Alimenti (C05): Verbal Fluency and Creative thinking: Comparing Semantic and Emotion categories and Network Dynamics

Bedia Vidua (C05): Investigating the Role of Personal Experience in Semantic Clustering: Implications for Creativity and Divergent Thinking

Dilara Balaban & Amelie Ewald (C01): Exploring Linguistic Creativity: Methods and Challenges of a Production Task.

Project C01

hintere Reihe v.l.n.r.: Joana Cholin, Lisa Hudalla, Amelie Ewald

vordere Reihe v.l.n.r: Jana Häussler, Dilara Balaban

Was uns an Sprache fasziniert. Nachlese zur VHS-Veranstaltung #5

2025-11-06T23:35:24+01:00

„Bielefeld – Liebefeld. Wie wir mühelos zu neuen Wörtern kommen“: Im Zentrum der ersten Veranstaltung des neuen Semesters und damit der fünften in unserer Reihe „Faszination Sprache“ stand am 6. Oktober der Beitrag der Psycholinguistik für die Bildung von kreativen Wortschöpfungen. Wie gewohnt, verständigten sich die Mitglieder des SFB, in diesem Fall Jana Häussler und Amelie Ewald vom Projekt C01, mit dem Publikum über das Thema anhand von Beispielen aus der aktuellen Forschungspraxis.

Dafür brachten sie Bilder von Gegenständen mit, für die es im Alltag keine etablierten Bezeichnungen gibt, wie der Schlitz für die Pfandmünze im Einkaufswagen, eine Pizza auf drei Beinen oder eine Maus mit Erdbeerfell. Für die Forscherinnen, war es sehr spannend zu verfolgen, wie sich Schritt für Schritt die Liste möglicher Bezeichnungen füllte: vom „Geldreinstecker“ und „Einkaufswagenchiphalter“ über den „Pizzahocker“, „Pizzatisch“ und „Esspausenhocker“ zur „Beerenmaus“ und dem „Erdbeersüßmäuschen“. Nach dem Input darüber, wie wichtig für erfolgreiche Wortbildungen der Anschluss an bereits vorhandene Einträge im mentalen Lexikon ist, wurden die Beispiele nochmals gemeinsam reflektiert. Schließlich erstellten die Teilnehmer*innen Hitlisten, wobei für sie zur Beurteilung neben der Originalität Kriterien der Eindeutigkeit und Verständlichkeit eine zentrale Rolle gespielt haben.

Unsere nächste Veranstaltung findet schon am Montag, den 10. November statt. Dann wird es darum gehen, „wie Gesten uns bei der Kommunikation unterstützen können, wenn die Worte schwerfallen“…

Mehr Infos dazu gibt es hier.

Zur Anmeldung für diesen Termin.

Fotos: ©Sascha Hermannski/Tim Rittig, SFB 1646

A05 and B02 at GENIALE 2025: Game Night with AI

2025-11-05T09:43:12+01:00

On 22 October, projects A05 and B02 of the CRC's Digital Linguistics Lab (DiliLab) and the Computational Linguistics Group (CLAUSE) took part in the "KI, ärgere Dich nicht!" Game Night, which was organized by the KI Akademie OWL as part of the Geniale Festival „KI und wir“ at the Wissenswerkstadt.The concept of the language games was developed by Clara Lachenmaier. PhD students Manar Ali and Omar Momen and research assistants Hannes Gröner, Helena Stahlschmidt, Raha Askari, Larissa Koch and Viktoria Rupprecht had developed various interactive games that invited participants to engage with language models through playful experiences.

Featured Games:

Game 1: AI Can't See?

In this visual puzzle game, players help a vision-language model locate a specific color grid using textual hints. This challenge explores the interplay between human creativity in giving clues and the model's limitations in processing visual and textual information. Game Designer and Presenter: Manar Ali (B02)

Game 2: Be Metaphoric!

Can you craft a statement so uniquely unconventional that no one has ever expressed it that way before? This game challenges the notion that creativity might be the one domain where humans will always surpass AI. Game Designer: Omar Momen (A05), Presenter: Victoria Rupprecht (B02)

Game 3: Codenames with a Twist

Based on the popular word association game, two teams exchange coded clues through their spymasters (e.g., "water 2" to hint at "shark" and "ice"). Our unique twist: one spymaster is actually a large language model. Players must not only win the game but also detect whether their clues came from a human or a machine. Game Designer: Hannes Gröner (CLAUSE), Presenter: Larissa Koch (B02)

Game 4: Do the Opposite!

Do-The-Opposite is a multiplayer, multi-round game where the players' individual creativity is tested by giving them instructions like "Talk like a baby" and asking them what they would do as the opposite. Players answer anonymously and, during a poll, are asked to vote for the most creative answer. Here, an answer generated by an LLM is also added to see whether the LLM can show more novel answers. The winner (a player or LLM) is the one with the most votes at the end. Game Designer: Raha Askari (CLAUSE), Presenter: Viktoria Rupprecht (B02)

The GENIALE was a very engaging experience for us! Visitors, particularly the younger participants, showed outstanding creativity and enthusiasm. The high number of attendees who enjoyed the activities made this an ideal opportunity to discuss language models with the public and conduct a pilot study, helping us refine our experimental design and optimize future data collection.

A report on the Game Night was published in the Westfalen-Blatt from 25^th Oktober 2025.

In the last month of Geniale, we have three more interesting events:

„Wissenshappen. Autor: ChatGPT?“ on 8^th November 2025, Projects: A05, B02, Ö
"Guck mal, wer das spricht" KI-generierte Podcasts im Selbstversuch on 13^th November 2025, Projects: B02, Ö
"Wer hat's geschrieben? Mensch oder Maschine?", Mitmach-Station until 29^th November 2025, Projects: A05, Ö

Team Members from left to right: Larissa Koch, Omar Momen, Raha Askari, Manar Ali, Helena Stahlschmidt

Conference Review: The 1st Summer School on Linguistic Creativity Hosted by the CRC 1646 in Bielefeld

2025-10-29T15:46:20+01:00

From July 21^st to 25th 2025, the First Summer School on Linguistic Creativity was hosted by the CRC 1646 „Linguistic Creativity in Communication“. International experts and participants came to Bielefeld University to discuss linguistic creativity in different fields, gain insight into various methods and engage in scientific exchange. We welcomed our external participants with a joint dinner at the Finca Bar Celona on Sunday (20.07.), before the Summer School started the next day.

The summer school program splits into workshops in the morning and working groups in the afternoon. On the first two days (Monday, 21.07, and Tuesday, 22.07.), the workshops were about the theoretic foundations of creativity in Syntax and Semantics/Pragmatics as well as the role of individual differences and social context for creative ideation.

Creativity in Syntax

Jutta Hartmann, speaker of the CRC, and Anke Himmelreich held the workshop on creativity in (minimalist) syntax. They gave an introduction into current assumptions and core concepts of the Minimalist framework, such as the operations Merge and Move, as well as an overview of the definitions of creativity within the CRC. The main research question that this theory-driven workshop aimed to answer was how such creative (morpho-)syntactic structures can be described in a fixed conventional system such as Generative Grammar and how analogy as a core mechanism can be used to explain different syntactic phenomena such as long-distance island extractions.

Individual Differences

This workshop was divided into three parts:

By giving a historical overview of the development of experimental approaches and reasonings Annett Jorschick, from Bielefeld University, showed how individual differences became increasingly important in experimental research, shifting the focus from purely observing group-level differences to individual personality traits. This was followed by considerations on the statistical modeling of individual differences and the challenges and pitfalls associated with it.

Nils Frederik Tolksdorf, from Heidelberg University, addressed individual differences in children, like linguistic abilities and temperamental shyness. We learned about the role it plays in children‘s creative and adaptive strategies in interactive language learning settings with humans and social robots.

The workshop was rounded off by Katalin Havas and Martin Wegrzyn, from Bielefeld University, with a focus on the creative process itself. In this hands-on session, we looked at the connection between verbal fluency behavior and creative products in the Divergent Association Task based on self-collected data. Other factors, such as IQ and personality traits, were also discussed.

Semantics/Pragmatics

Bob van Tiel, from Radboud University Nijmegen, was the lecturer for the Semantics/Pragmatics workshop. After he made the basics of pragmatics accessible to everyone, he engaged with us in a discussion about lexical scales. We learned that such lexical scales are everywhere: While the prominent example some vs. all is well researched, Bob drew our attention to other scales like between big and small, or small and tiny. On the basis of different experiments in the field, we discussed theoretical insights and methodological considerations.

The workshops on Wednesday, 23.07., and Thursday, 24.07., were about methods for investigating linguistic creativity, covering Corpus Analysis, Acceptability Judgements and Multimodal Creativity Research.

Corpus Analysis

In the workshop given by Nils Reiter, from University of Cologne, the participants examined annotation as a research practice, beginning with how to design effective annotation guidelines from scratch. The workshop gave an opportunity to discuss strategies for evaluating annotations, to explore what these evaluations reveal, as well as to consider different ways of structuring annotation rounds in practice. The participants also looked at how machine learning models can be used to support and accelerate the annotation process, with opportunities to explore their potential in practice.

Acceptability Judgments

The workshop led by Thomas Weskott, from University of Göttingen, focused on the methodologies and considerations related to acceptability judgments in linguistic research. Key discussion topics included the debate on using continuous versus discrete measurement scales and the role of benchmarking. The workshop also provided knowledge on applying this method in controlled experimental settings and demonstrated analytical tools for evaluating and visualizing judgment data in R. A particular emphasis was placed on imparting descriptive and inferential statistical knowledge for analyzing linguistic intuition.

Multimodal Creativity Research

Hans Rutger Bosker, from Radboud University Nijmegen, led a two-day hands-on workshop on multimodal creativity research. After presenting insights from experiments conducted in the SPEAC Lab, he invited the participants to design their own mini-experiment, run it online, and analyze the results in the second session. By creating our own experiment and discussing the outcomes, we gained a deeper understanding of multimodality and its potential in research.

On Friday, 24.07., two parallel workshops on fundamentals took place:

Verbal protocols

Lena Heine, from Ruhr University Bochum, introduced the methodology of verbal protocols. She guided us through the underlying framework of the interdependency of thoughts and language and illustrated the strengths of offline methods in cognition research as well as the challenges in the interpretation of verbal protocols by using examples from an educational context.

Visualization in R

Elen Le Foll, from University of Cologne, gave an introduction to data visualization in the R programming language. She highlighted key features and demonstrated useful functions and packages for data visualization. During the session, participants not only reflected on data visualization in general but also had the opportunity to experiment with R themselves.

In the afternoon of each day, there were working groups with different talks from participants, who took the opportunity to present their research to each other. External participants talked for example about creative cursings in Tunesian Arabic and orthographical and lexical creativity in Luxembourgish youth chats. Internal talks presented current work within the different CRC projects (see below for all presentations).

Of course, the summer school was accompanied by social events. For example, we played board games together and even visited a medieval market.

All in all, it was an all-round successful summer school that has whetted our appetite for the next one. We are looking forward to hosting the DGfS Summer School in Bielefeld next year. So save the date for it: August 10th–21st!

External and internal talks were:

Mareike Hartmann (C03): Exploring Linguistic Creativity in Naming and Interaction: Creative Expressions as a Pathway to Communicative Success in Individuals with Neurological Language and Communication Disorders

Lotta Heidemann (C03): Patterns of Multimodal Creativity in People with Neurogenic Speech and Language disorders

Afef Labben (University of Tunis): Creative Impoliteness: An Interdisciplinary Study of Cursing in Tunisian Arabic

Eleonore Laubenstein (A04): To Agree or not to Agree, that is the Conflict: Exploring Verbal Agreement with Disjunctive Subjects in German

Melissa Mujkić (University of Luxembourg): Orthographical and Lexical Creativity in Luxembourgish Youth Chats

Tijana Šuković (University of Belgrade): How Creative and Context-Dependent are Non-Lexicalized Occurrences with Proper Nouns in English?

Jiaxuan Xie (University of Aberdeen): Linguistic Creativity and Gendered Pragmatic Misalignment in Autistic Women: A Cross-Cultural Sociolinguistic Analysis

The Program of the Summer School

The Research Training Group (RTG)

The DGfS-Summer School 2026 in Bielefeld

Conference Review: Project A03 at the SALA-39 in Poznań

2025-10-15T12:07:54+02:00

South Asian Languages Analysis Roundtable (SALA) is a prestigious annual conference for linguistics in this field. This conference was founded by Prof. Hans Henrich Hock in 1978 in the Linguistic Institute of the Linguistic Society of America at the University of Illinois, Urbana-Champaign. Since then, SALA is hosted each year at different locations around the world.

This year, SALA 39 was organized by the Institute of Oriental Studies at the Faculty of Modern Languages and Literatures, Adam Mickiewicz University at Collegium Novum in Poznań, Poland from 29 September to 1 October 2025.

The conference covered a wide range of topics in linguistics, including (but not limited to) phonetics, morphosyntax, typology, and semantics. The keynote lecture by Prof. Miriam Butt from the University of Konstanz on Modern Indo-Aryan Case Marking set the tone for the event and stimulated engaging discussions throughout the sessions.

On the third day, Shravani Patil and Farhat Jabeen from project A03 presented their talk titled “Variability in (Non-)Canonical Word Order and Focus Marking in Hindi/Urdu”. The presentation examined the interface between word order and information structure - an important area of research within project A03. It demonstrated how new information focus is marked in Hindi/Urdu and highlighted variability in speakers’ word order preferences in marking focus.

The talk was followed by a brief discussion session. The feedback proved to be important not only in terms of further development of this study but also served as an interesting discussion point on the use of differential strategies to mark different information structure phenomena.

Link to the Abstracts

Conference Review: Project A01 at the 17th International Conference on the Structure of Hungarian (ICSH17) in Debrecen

2025-10-09T21:34:15+02:00

The ICSH is a biennial conference concerning any aspect of the Hungarian language. The host country alternates between a Hungarian and a non-Hungarian location every time, and the 17^th conference was held at the University of Debrecen, Hungary this year (11-12 June). The invited speakers were Ora Matushansky (SFL, Centre National de la Recherche Scientifique/Université Paris 8) and Markus Pöchtrager (Universität Wien). The conference was followed by a thematic workshop on “New Insights on Demonstratives” (13 June), with Elsi Kaiser (University of Southern California) being the invited speaker.

The conference comprised of fifteen talks which targeted a range of synchronic/diachronic phenomena from different angles (via experiments and various theoretical frameworks). The final two presentations were given by A01’s PhD student Szilvia Daczó and PI András Bárány, respectively. Since there were four submissions related to object agreement in Hungarian, a panel discussion was held to round up the main related issues and to close off the conference. The panel was led by the renowned Katalin É. Kiss (HUN-REN, MTA, ELTE), and a few thought-provoking lightning talks were given by Anikó Lipták, György Rákosi, András Bárány, and Szilvia Daczó.

Szilvia’s talk on long-distance agreement with intransitive matrix verbs in present-day Hungarian detailed the work she has done so far as part of the project A01 that investigates morpho-syntactic creativity. Regular object-verb agreement is a well-known feature of Hungarian, yet – with the exception of Bárány 2020 – not many have explored those verb-infinitive-object structures where object agreement surfaces on a (non-object-selecting) intransitive verb due to the presence of a definite object selected by a transitive infinitive. Szilvia presented the results of her corpus work, discussing tense, person-number, and constituent order distributions, as well as verb-specific variation. These raised questions about acceptability among native speakers of Hungarian, a pilot test of which was the topic of the remainder of her talk. The results of the experiment indicate that this structure is moderately grammatical, with factors such as constituent order and analogy from the equivalent structure with canonical transitive matrix verbs having an additional – though not always favourable – effect on acceptability judgements. The audience then provided useful insights and questions for future work.

Thereafter, Szilvia brought up a secondary yet intriguing aspect to her dataset during the fast-paced panel session. She felt very honoured to engage with those most knowledgeable scholars who have majorly contributed to our understanding of object agreement in Hungarian.

The final talk was given by András, who contrasted the two main views Hungarian object markers – namely, whether they are in nature agreement markers or clitics. Applying cross-linguistically established morphological, semantic, and syntactic diagnostics for identifying clitics and agreement markers, he argues that Hungarian has morphologically bound agreement markers, unlike languages such as Inuktitut, Amharic, and multiple descendants of Indo-European.

Regarding the panel, András’ contribution dealt with one particular form in the Hungarian object agreement which has been characterised as special in the literature. András argued that a change of perspective on Hungarian object agreement makes the “special” form -lak/-lek a fully regular part of the overall system.

Reference:

Bárány, A. 2020. Hosszú távú egyeztetés intranzitív igékkel a magyarban [Long-distance agreement with intransitive verbs in Hungarian]. ÁNyT 32. 55–68

See also: The Programme of the ICSH17, Abstracts by Szilvia Daczó and András Bárány

The Poster of ICSH17

Was uns an Sprache fasziniert. Nachlese zur VHS-Veranstaltung #4

2025-10-08T23:51:24+02:00

Wie kreativ Kinder im Spracherwerb sind, war Thema der vierten Veranstaltung in unserer VHS-Reihe „Faszination Sprache“ am 16. Juni 2025, zu der das Team des Projekts C04 in Austausch mit sprachinteressierten Bürger*innen kam.

Das Publikum konnte an diesem kurzweiligen und informativen Abend die Erfahrung machen, dass der Kreativität bei der Neuschöpfung von Wörtern keine Grenzen gesetzt sind. Während Kinder im Vorschulalter wahre Sprachgenies sind, begleitet uns die Suche nach Bezeichnungen für neue Gegenstände oder Phänomene ein Leben lang. Wie Alexandra Filimonova, Denise Gajda, Martina Hielscher-Fastabend und Julia Settinieri näher ausführten, gibt es sehr unterschiedliche Strategien, zu neuen Wörtern zu kommen. Im Deutschen führen die zusammengesetzten Wörter die Hitliste an, oft auch verbunden mit Funktionsbeschreibungen, wie im Fall der „Gießkanne“. Wer das Glück hat, mehrsprachig aufzuwachsen, ist eindeutig im Vorteil, weil man dann Code-Switching betreiben kann. Dabei hängt es von der kulturellen Bedeutung des zu benennen Objekts in der jeweiligen Sprache ab, ob etwa aus dem „schafigen Käse“ ein „Schafskäse“ wird, wie im Deutschen oder ein eigenes Wort wie der „Feta“ z. B. im Griechischen.

Die Forscherinneren verstanden ihre Ausführungen auch als Beitrag zu mehr Kreativität in Sprachförderung und –therapie. Besonders interessant waren für sie Beispiele aus dem Publikum. So erfuhren sie u.a. von der im Alltag noch nicht gebräuchlichen Bezeichnung „Elter“ für ein Elternteil, wie man inzwischen in der Biologie und Soziologie sagt.

v.l.n.r: Martina Hielscher-Festabend, Denise Gajda,

Alexandra Filimonova, Julia Settinieri

Rückblick: Talkaoke in der Wissenswerkstadt: „Wie sähe eine Bielefelder KI aus?“

2025-10-01T23:36:35+02:00

Von ChatGPT über KI-generierte Schlumpf-Avatare bis hin zur „Bielefelder KI“. Beim Eröffnungswochenende der GENIALE zu „KI und wir“ in der Wissenswerkstadt wurde sieben Stunden lang diskutiert, gelacht und philosophiert. Das Projekt Ö des SFB 1646 „Sprachliche Kreativität in der Kommunikation“ lud am 6. September 2025 gemeinsam mit dem Londoner Künstlerkollektiv The People Speak zur Mitmach-Talkshow Talkaoke ein, um sich mit den GENIALE-Besucher*innen darüber auszutauschen, was sie bei diesem Thema bewegt.

Was ist Talkaoke?

Talkaoke funktioniert gleichzeitig ganz simpel und ziemlich verrückt: in der Mitte ein pink leuchtender, donut-förmiger Tisch und eine Moderation mit Mikrofon. Darum herum Plätze für alle, die reden wollen. Die Besonderheit: Nicht die Moderation gibt die Themen vor, sondern die Runde.

„KI und wir“ – Darüber wurde gesprochen

Am 1. Geburtstag der Wissenswerkstadt öffnete die Wissenswerkstadt die Türen für das Programm der GENIALE. Die GENIALE findet in ihrem ersten Jahr in der Wissenswerkstadt statt und hat „KI und wir“ als Motto. Zu diesem Anlass wurde bei Talkaoke vor allem darüber diskutiert, wie KI unser Leben verändert. In der siebenstündigen Gesprächsreise mit wechselnden Teilnehmenden kamen ganz unterschiedliche Themen auf den Tisch:

Eine Frage, die immer wieder auftauchte und die sofort zum Philosophieren anregte: Was ist eigentlich KI? Klar, eine „Künstliche Intelligenz“ – aber was macht sie intelligent? Was weiß eine KI, und gibt es dann auch eine „unkünstliche“ Intelligenz? Schnell ging es ins Praktische über: ChatGPT kannten alle in der Runde und die meisten haben es bereits genutzt, so etwa für das Schreiben lästiger Mails oder, um Anregungen zu bekommen. Andere nennen ChatGPT ihren „ganz persönlichen Assistenten“ und geben dem System sogar einen Namen. Gleichzeitig wurde Skepsis angesprochen: Wo gehen die Daten hin, die wir eingeben? Wie viel sollten wir über uns preisgeben? Auch der hohe Stromverbrauch von KI-Nutzung war Thema und ob wir freundlich zu ChatGPT sein sollten, dabei auch „Bitte“ und „Danke“ sagen – oder verbrauchen wir damit nur unnötig Energie?

Die Gespräche am Talkaoke-Tisch mit immer wieder wechselnden Teilnehmenden nahmen Abbiegungen in ganz verschiedene Richtungen: Jugendliche teilten mit, dass sie eine digitale „Schlumpf“-Version von sich selbst erstellten. Auch über KI im Verkehr wurde gesprochen: Wie sähe eine Welt aus, in der alle Fahrzeuge KI-gesteuert sind? Sollte man sie vom normalen Verkehr trennen? Würde optimierter Verkehr mehr Platz für Radfahrer und Fußgänger schaffen? Oder lassen wir uns in Zukunft einfach alle von Robo-Taxis kutschieren? Und an einem Punkt gestaltete die Runde eine KI, die aus Bielefeld kommt. Die Vorschläge: deutschsprachig, sachlich-freundlich, datenschutzbewusst und mit kurzen, knackigen Antworten.

Insgesamt zeigten die Gespräche vor allem eins: Wir machen alle ganz verschiedene Erfahrungen mit Künstlicher Intelligenz. Und: Es lohnt sich, darüber zu reden. Von Schüler*innen bis Senior*innen, von KI-Expert*innen bis Neueinsteiger*innen: Alle kamen zu Wort und konnten ihre eigenen Erfahrungen, Fragen und auch Sorgen teilen.

Der wissenschaftliche Blick auf Talkaoke

Das Projekt Ö des SFB wird die aufgezeichneten Gespräche nun wissenschaftlich auswerten: Welche Themen kamen auf, wie wurde über KI diskutiert und wie eignet sich Talkaoke als partizipative Talkshow, um Beteiligung in Diskussionen zu wissenschaftssensiblen Themen zu ermöglichen?

Über The People Speak

Das Format Talkaoke wurde von The People Speak entwickelt. Das Künstler*innen-Kollektiv entwickelt spielerische Formate, die Menschen ins Gespräch bringen. Ihr Motto lautet „Tools for the world to take over itself“ – also Werkzeuge, mit denen die Welt selbst das Ruder übernehmen kann. Bereits im August vergangenen Jahres waren The People Speak zu Gast beim SFB in Bielefeld und führten mehrere Talkaoke-Events durch. Zudem gab das Team im Anschluss auch dieses Mal einen Workshop, in Wissenschaftler*innen des SFB bei der Entwicklung neuer partizipativer Formate für ihre Forschung unterstützt wurden.

Der SFB 1646 auf der GENIALE

Auch im 2. und 3. Monat der GENIALE ist der SFB 1646 wieder mit eigenen Veranstaltungsformaten beteiligt. Hier ist ein Überblick:

"Wer hat's geschrieben? Mensch oder Maschine?", Mitmachstation ab 1. Oktober 2025 (Projekte A05 u. Ö)
KI-Forschung erleben. Labführung im CITEC, u.a. mit dem Projekt C02 am 2. Oktober 2025
"Guck mal, wer das spricht" KI-generierte Podcasts im Selbstversuch, Workshops am 15. und 24. Oktober sowie am 13. November 2025 (Projekte B02 u. Ö)
Wissenshappen. Autor: ChatGPT? am 8. November 2025 (Projekte A05, B02, Ö

Was uns an Sprache fasziniert. Nachlese zur VHS-Veranstaltung #3

2025-09-29T22:00:04+02:00

„Wovon hängt es ab, ob kreativer Sprachgebrauch gelingt?“: Diesem Thema war unsere dritte Veranstaltung in der VHS-Reihe „Faszination Sprache“ gewidmet, zu der das Team des Projekts A06 am 12. Mai 2025 in Austausch mit sprachinteressierten Bürger*innen kam.

Was ihn an Sprache fasziniert, machte Dr. Said Sahel gleich zu Beginn seiner Einführung ins Thema deutlich: Obwohl es uns so vorkommt, als würde sich die Sprache ohne unser Mittun wandeln, beeinflussen wir durch die Art und Weise unserer Kommunikation Sprachwandel aktiv mit. Durch Beispiele gegenwärtig ablaufenden Sprachwandels wird die Variabilität des Deutschen besonders greifbar. Anhand eines Ratings von Konstruktionen wie „Anna ist am Kochen, als ihre Mutter anruft“ und „Ich bin den Film am Kennen.“ konnte das Publikum sein Sprachempfinden walten lassen und individuell beurteilen, inwieweit ihm solche Formulierungen hier und heute als üblich erscheinen.

Wie Prof. Dr. Tanja Ackermann danach ausführte, gehen derartige „am-Progressiv“-Konstruktionen bis ins 15./16. Jahrhundert zurück und werden allerdings heute innerhalb des deutschen Sprachgebiets unterschiedlich bewertet. Der gemeinsame Blick in den Atlas zur deutschen Alltagssprache verriet: Im Rheinland sind Sätze wie „Ich bin am Schlafen.“ oder „Ich bin die Uhr am Reparieren.“ häufiger zu finden als in Westfalen oder Niedersachsen. So spiegeln sich unterschiedliche Phasen eines Sprachwandelprozesses auch in geographischen Unterschieden. Wie das Live-Experiment zur Formulierung „Ich fahre E/essen.“ demonstrierte, kann die eigene Sprachbiographie darüber mitentscheiden, wie die Äußerung interpretiert wird.

Weitere Beispiele verdeutlichten an diesem unterhaltsamen und lehrreichen Abend, wovon es abhängen kann, ob zunächst ungewöhnliche Formulierungen sich einbürgern und irgendwann als normal empfunden werden. Am Schluss war es sogar möglich, einen Eindruck davon zu bekommen, wie die historische Sprachwissenschaft heutzutage vorgeht. In der Forschung können einzelne Wörter und Wendungen in digitalisierten Textkorpora recherchiert und ihre sprachlichen Eigenschaften bestimmt werden. So kann der Blick in die Geschichte auch dabei helfen, gegenwärtig ablaufenden Sprachwandel zu verstehen und einzuordnen.

Neben den beiden oben genannten Forscher*innen waren Dr. Désirée Kleineberg und Rebekka von Legat von A06 sowie das Projekt Ö an der Entwicklung der Veranstaltung beteiligt.

Die Veranstaltungsreihe "Faszination Sprache" in der VHS Bielefeld wird im Wintersemester 2025/26 mit weiteren Themen und Projektgruppen fortgesetzt.

Projekt A6:

hintere Reihe v.l.n.r.: Barbara Job, Tanja Ackermann, Said Sahel

vordere Reihe v.l.n.r.: Rebekka von Legat, Désirée Kleineberg

Review: Ákos Buza (A03) on his Research Stay at the Phonetics Research Group in Budapest

2025-09-28T18:11:01+02:00

From 15^th to 17^th of July, Ákos Buza, PhD student at Project A03, had the opportunity to visit the Phonetics Research Group at the Institute for General and Hungarian Linguistics of the Hungarian Research Centre for Linguistics (NYTK) in Budapest. During his time there, he recorded two speakers whose voices will serve as stimuli for his research, to help with the investigation into mismatches at the intersection of syntax, information structure, and prosody. The colleagues at the institute were incredibly helpful in setting up the phonetics studio to ensure Ákos captured high-quality audio.

This visit brought his project a step closer to gathering the data it needs, while also helping him to a better understanding of the prosodic patterns in his target sentences. In addition to this, Ákos was able to contribute to one of Budapest`s ongoing experiments by being recorded himself.

The visit was another step in the ongoing productive collaboration between Dr. Katalin Mády, head of the Research Group of Phonetics in Budapest, and the A03 project.

The Hungarian Research Center of Linguistics in Budapest

Control-room setup at the Hungarian Research Center for Linguistics in Budapest

Conference Review: The CRC 1646 hosted the GAKT. 9th International Workshop on German Linguistics between Cologne and Tokyo in Bielefeld on “Creativity and Interpretation”

2025-09-24T23:40:56+02:00

From 29 to 30 April the GAKT-Workshop (Germanistische Linguistik zwischen Köln und Tokio) took place for the first time on the premises of the CRC 1646 at Bielefeld University, which is why this event became GAKT-BI this year. The workshop was also a collaboration of two linguistic CRCs in Germany. Arndt Riester from the CRC 1646 Linguistic Creativity in Communication at Bielefeld University, Klaus von Heusinger from the CRC 1252 Prominence in Language at Cologne University and Yoshiki Mori from the University of Tokyo organised the event.

As part of the workshop, thirteen doctoral students and other researchers from Bielefeld, Cologne and Japan had the opportunity to present their work on the topic of “Creativity and Interpretation” to a broader audience. Selected linguistic phenomena in Japanese and German were analysed in terms of syntax, semantics and pragmatics, while several CRC members took the role of commentators. Fitting the collaborative spirit of the GAKT-workshop series some of the presentations were the result of long-lasting cooperations between Japan and Germany. In total, there were ten talks.

For example, Katsumasa Ito (Osaka Metropolitan University) and Heiko Seeliger (CRC 1252, projekt A06, C10) jointly presented their thoughts on the semantics and pragmatics of the German expressions "Und ob!" and "Als ob!". They explained, for example, why the two expressions mark the obviousness of the truth or, respectively, the falsity of a statement. Nori Hayashi (University of Tokyo) presented his ideas on the “Polarity asymmetry under possibility modality and attention pragmatics”. He is currently a guest researcher in CRC 1646 project B03.

Several members of the CRC 1646 participated in the workshop, either as presenters or as commentators. Eleonore Laubenstein and Robert Lange, members of project A04 in the CRC 1646 were commented by Anke Himmelreich (CRC 1646, associated member in project B03). In their joint talk titled “To agree or not to agree. Kongruenzkonflikte bei Disjunktionen im Deutschen“, they investigated the creative potential of disjunctive subjects in sentences such as “Du oder ich mach-en/-st/-e das”. Fabian Zöfelt, another member of the CRC 1646 (project A01) presented his work with the title “Wenn wir es nur wüssten. Syntaktische Ambiguität von und Extraktion aus wenn-Sätzen”. Among other things he discussed the ambiguos syntactic status of some “wenn”-clauses, followed by a comment by Jutta Hartmann (CRC 1646, project A01, B03). Mariya Hristova (project B04) presented her work on creative multiword expressions in doctor-patient conversations. She demonstrated the processes involved when negotating the meaning and usability of creative expressions with the example “durcheinanderer Mensch”. Deeply tied to the topic of linguistic creativity, her presentation was fittingly commented by another member of the CRC 1646 (Torgrim Solstad, project B01)

The empirical study and language comparison by Timo Buchholz (CRC 1252) were commented by Jana Häussler (CRC 1646, A04). His talk with the title “Effects of subordination on anaphor resolution in Japanese and German” dealt with the the extent to which syntactic, punctuational and discourse-structural characteristics of sentences influence how anaphors are interpreted – and whether there are differences between Japanese and German in this respect. As another example of the scientific exchange between Bielefeld and Tokyo, two of the organisers gave a joint presentation. Yoshiki Mori, cooperation partner of project B03, and Arndt Riester presented how different yes-no questions are formulated in Japanese and which contextual factors influence their acceptability.

On the evening of the first day of the workshop, a tribute was paid to the recently retired Prof. Dr Yoshiki Mori, who has been committed to German-Japanese exchange in German linguistics for over 40 years, by way of organising conferences, inviting German academics to Japan and serving on the international advisory board of the IDS (Institut für Deutsche Sprache) Mannheim. All in all, the 9^th outing of the GAKT-Workshop series proved to be a great opportunity for young researchers to build and strengthen international connections between Japan and Germany.

v.l.n.r.: Yoshiki Mori (University of Tokyo), Arndt Riester (CRC 1646, B03),

Rückblick: Désirée Kleineberg (A06) im Tübinger Oberseminar zu Gast

2025-09-22T13:16:42+02:00

Am 10. Juli 2025 stellte Désirée Kleineberg Teile ihrer Forschung fürs Projekt A06 im Rahmen des Oberseminars „Aktuelle Themen der romanischen Sprachwissenschaft“ an der Universität Tübingen vor. In ihrem Vortrag zum „Schicksal gescheiterter Kreativität“ präsentierte sie eine diachrone Korpusstudie der französischen Verbalperiphrase in ihrer Entwicklung zwischen dem 17. und 21. Jahrhundert. Im Zentrum des Vortrags stand die Frage, welche sprachlichen Veränderungen diese periphrastische Konstruktion durchgemacht hat. Nachdem sie zu Beginn des 17. Jahrhunderts plötzlich fast vollständig aus dem Sprachgebrauch verschwunden war, stellte sie ein Beispiel für langfristig nichterfolgreiche Kreativität dar. Ihre Analysen zeigten jedoch, dass die Konstruktion in einer „grammatischen Nische“ überlebt und für neue, wiederum kreative Ausdrucksmöglichkeiten der Verlaufsform Platz gemacht hat.

Im Anschluss ans Kolloquium hatte Désirée die Möglichkeit, am Tübinger Workshop „Correspondances familières francophones (XVIIe–XXe siècles)“ teilzunehmen. Die Beiträge in der von Rembert Eufe, Katharina Fezer und Inga Hennecke (Romanisches Seminar, Center for Francophone Worlds) organisierten Veranstaltung demonstrierten eindrucksvoll die Wichtigkeit von bisher oftmals wenig beachteten Quellen für die historische Sprachwissenschaft und deren Variabilität zwischen konventionalisierten und kreativen Schreibpraktiken in französischen Privatkorrespondenzen aus dem 17.-20. Jahrhundert.