SFB 1646

My research explores linguistic creativity via analogy from a morphosyntactic perspective, through a case study on Hungarian long-distance object-verb agreement with intransitive matrix verbs (iLDA). This is an uncanonical structure, and has largely been neglected in previous work. My PhD will offer insight into the factors that motivate the use and influence the comprehension of iLDA, besides furthering its theoretical analysis. I will empirically test iLDA via acceptability judgments and self-paced reading. Thereby, I contrast off-line and on-line processing of an uncanonical phenomenon, and measure it against canonical long-distance agreement, which is conceptually the source analogue to iLDA. Currently, I am running a pilot self-paced reading experiment, where all participants read iLDA (besides fillers and distractors) in the second block, but differ in what they read in the first block. In block 1, group A reads how intransitive verbs usually inflect in this configuration (providing a morpho-syntactic baseline), group B reads canonical LDA (potentially facilitating analogy), group C reads iLDA (providing an adaptation baseline). Evidence for analogy seems dubious so far, when reviewing corpus, acceptability, and self-paced reading work.

 

Status: ongoing 

My research aims to investigate how co-speech gesture supports both speakers and listeners in inventing and using new linguistic constructions, for example to solve challenging communication situations. Therefor, my project will examine in studies with human speakers how co-creative gestures can be characterized and generated by trained models. The long-term goal is to develop a cognitive-computational model of multimodal speech-gesture use.

 

Status: ongoing

I'm currently working on a systematic analysis of the interplay between linguistic and communicative skills, cognitive processing, and verbal creativity in individuals with language and communication disorders resulting from a brain lesion (stroke). My PhD project examines the extent to which persons with aphasia (PWA), persons with cognitive-communicative deficits (PCCD), and neurologically unimpaired control persons (NUC) can be linguistically creative (novelty or originality of linguistic expressions) and how the creative use of language in PWA and PCCD influences the success of communication in social interaction (effectiveness of linguistic expressions). For this purpose, two experimental tasks were designed: (1) the oral naming of objects (using compound words) and actions (using verbs) and (2) the interactive description of compound nouns based on the taboo-paradigm. Both tasks place high demands on the retrieval and processing of German compound words and on communicative skills. They therefore require creative language use, not only in PWA and PCCD but also in NUC, enabling the comparison of approaches used by persons from the different groups.

 

Status: ongoing

I am currently working on an analysis of multimodal creativity in people with neurogenic language and communication disorders. This analysis includes the description and coding of creativity in speech replacing and speech accompanying gestures, and is concerned with the development of a novel coding scheme that aims to capture key aspects of creative patterns in gesture production. Furthermore, the investigation will encompass the potential significance of relevant underlying (cognitive) abilities, including general creativity and executive functions. By a comparison of neurologically unimpaired controls, people with aphasia (due to a left hemispheric lesion) and people with cognitive or communicative deficits (due to a right hemisphere lesion), we investigate the interplay between linguistic and communicative abilities, cognitive processing and patterns of multimodal creativity.

 

Status: ongoing

 

 

 

 

 

The aim of this dissertation project is to investigate linguistic creativity, with a particular focus on semantically creative expressions such as metaphors and similes. In a first step, these expressions are classified according to their source domains. Building on this, the project explores whether there is a correlation between specific source domains and the degree of novelty of these expressions. While the first part of the dissertation provides a systematic overview, the second part focuses on selected conversation sequences that contain creative expressions. It examines how their meaning is negotiated between conversational partners. To account for this process, the project applies formal linguistic modelling, drawing on an extended discourse model that captures the role of indirectly communicated information, which is crucial for the interpretation of creative language.

 

Status: ongoing

My dissertation covers two different topics. From a computational linguistics perspective, I am interested in the discourse capabilities of LLMs and how they manifest in the way they generate language. I investigate this mostly in the domain of causal language use through phenomena like Implicit Causality. On the other hand, I am studying how to leverage citizen science paradigms for linguistic research on the web. This part of my work focuses on both the technical and methodological aspects of developing the infrastructure necessary to conduct web-based citizen science research and how to design and implement such research programs.

 

Status: ongoing

I am currently analyzing my corpus data. This consists primarily of conversation data from 20 native German speakers, who converse with other native German speakers in one set of recordings and with non-native speakers in another. I am examining various parameters to see if there are differences in speech behavior, for example, in terms of linguistic complexity.

 

Status: ongoing

I'm currently working on agreement with disjunctive subjects in German as an example for grammatical conflicts. While agreement with conjoined subjects is done using a plural morpheme on the finite verb, there are multiple ways for the verb to agree with disjoined subjects in German. This shows that there is no definitive rule for speakers to adhere to when forming sentences with disjoined subjects. Instead, speakers use other tools in their grammatic arsenal to determine what form the verb should take, either using a plural analogous to conjoined subjects, or an ad hoc choice based on other factors. Determining what these other factors are and how they influence agreement with disjunctive subjects is the aim of this research. 

 

Status: ongoing

I'm currently working on the interactive dynamics of creative utterances in B04's recorded doctor-patient interactions. Using the means of (Longitudinal) Conversation Analysis, I investigate what happens in interaction when the interlocutors use creative language: Why do the patients use creative language? How does the doctor react to their creativity? In which cases do the creative expressions re-appear later in the conversation, so that an observable routinization is taking place?

 

Status: ongoing

I'm currently working on setting clear tasks of processing creative linguistic signs, this includes metaphor novelty scoring, general linguistic originality and meaningfulness ratings. I use both the probability distributions and generated responses of LLMs during these tasks to probe their hidden representations, attention weights and representations, and MLP networks. My research tries to approach the question of whether LLMs perform certain interal algorithms/mechanisms during processing creative signs. 

 

Status: ongoing

Focus marking is utilized differently in different languages. In Project A03, we aim to investigate creativity in the interpretation of mismatching input at the interface of (morpho)syntax, information structure, and prosody. We work different languages like German, Hindi/Urdu, and Hungarian as they use varied cues to mark focus. We are specifically interested in listener based data. I am currently working on investigating focus marking in Hindi/Urdu by conducting experiments and via corpus study. Some of our common methodologies include production studies, picture selection tasks, rating experiments. 

 

Status: ongoing

My PhD project forms part of subproject Ö of CRC 1646 at the interface of Science Communication and Linguistics and explores how participatory approaches can be integrated into linguistic research on creativity. 
I will analyse which engagement formats are suitable for which type of research question, process and target audience. I will also investigate the potential benefits of engaging the public in linguistics, such as new forms of knowledge production. To achieve this, I use qualitative data from the preparation, realisation and follow-up reflection phases of formats to identify overarching principles for public engagement in linguistics. Ultimately, I aim to develop a framework that will support researchers in designing and evaluating participatory linguistic studies, including practical conditions, challenges and best practices.

 

Status: ongoing

I am working on linguistic creativity in literary and non-literary text from a Computational Literary Studies perspective. My project aims to contribute to a deeper understanding of linguistic creativity by identifying the specific conditions under which linguistic signs are perceived as creative. In a textual domain like literary text, creativity is never only a textual property at the level of the sign. Instead, it emerges in usage in an interaction between linguistic form, genre, and reader response. I am interested in how these components interact when readers, who themselves have certain expectations and prior experiences, perceive creativity. I develop and evaluate empirical and computational metrics for automatically assessing these components, bridging conceptual and empirical approaches to creativity.

 

Status: ongoing

 

My current research interest is in open texture as a philosophical problem of linguistic indeterminacy. I work to contrast this kind of indeterminacy in particular with that of vagueness, and to investigate how open texture is realized in particular semantic structures. My work in progress focuses on the possibility of terms featuring prototype semantic structure, and how this structure realizes the indeterminacy of open texture. I maintain a long-term view to the problem of connecting open texture to linguistic creativity.

 

Status: ongoing

I am currently investigating creative communication in four- and five-year-old children when they encounter lexical gaps. My research focuses on the multimodal strategies children use in such situations, particularly creative gestures and other non-verbal resources that support communication and facilitate understanding. In addition, I explore how individual differences in shyness influence children's communicative behavior. I am also interested in how situational resources and interactional contexts shape children's responses to communicative challenges and the strategies they employ.

 

Status: ongoing

In my project, I investigate the role of analogy in the creation of innovative linguistic expressions. More specifically, I work on long-distance dependencies involving complement-like *wenn*-clauses in German. These embedded domains are generally considered to be syntactic adjuncts. Consequently, complement-like *wenn*-clauses should constitute strong islands for extraction. In my current experiments, I have, however, found that extraction from these clause types appears to be possible to some extent for a subset of participants. Interestingly, participants who moderately accept these dependencies also allow extraction from factive *dass*-complements. I propose that this pattern can be explained by expansion in the sense of analogical reasoning. I am also interested in whether such creative constructions become more acceptable with repeated exposure, that is, whether extraction from complement-like *wenn*-clauses can be primed.

 

Status: ongoing

My habilitation project focuses on psycholinguistic and cognitively plausible modelling approaches to human speech production and perception. This work combines Computational Linguistics, Psycholinguistics and Experimental Phonetics by applying state-of-the-art computational modelling with Artificial Neural Networks / Large Language Models (LLMs) to the interactional dynamics and ubiquitous variability in spoken language. In particular, I focus on non-canonical speech production data shaped by social, psychological and cognitive factors.

 

Status: ongoing

In my postdoctoral research project, I examine the contextual factors underlying the diffusion and decline of Romance verbal periphrases involving a motion verb and a gerund/V-ant. While Romance languages – I concentrate on French, Spanish, and Italian – generally all possess verbal periphrases with the corresponding structure and a durative function (e.g., Spanish ir haciendo), they are characterized by varying degrees of productivity in their contemporary forms. In particular, no satisfactory, empirically grounded explanation has yet been provided for the decline of those constructions that have an extremely low frequency today and exhibit strong combinatorial restrictions (especially in French). As part of my postdoctoral research, I aim to close this research gap by analysing not only the immediate co-text but also broader contextual levels. A particular focus here is on considering discourse traditions as “carriers” of the diffusion, but also of the decline, of linguistic innovations. Methodologically, I combine diachronic quantitative-qualitative corpus analyses with the analysis of additional sources and metalinguistic documents. My primary focus is on the Early Modern Period, which appears to have been decisive for the further development of the periphrases I am investigating.

 

Status: ongoing

I'm currently working on a frame-based approach to the interpretation of complement coercion constructions. I aim to modell the mechanisms and processing steps involved in enriched composion at the syntax-semantics-pragmatics interface, considering the possibilities of creative interpretation as well as its limits based on observed data. I am also involved in the project's empirical work, mainly the production studies outlined in our Work Package 1.

 

Status: ongoing