Automatic and controlled processing in conflict tasks

(Project of the Baden-Württemberg foundation)

We are constantly required to select and process task-relevant information in environments overloaded with potentially distracting information. One central goal of cognitive psychology is to uncover the interplay of the unobservable mental processes that allow goal-directed behavior. One useful approach is to investigate task performance (reaction times) in simple perceptual decision-making tasks under the presence of multiple sources of information, both relevant and irrelevant („conflict tasks“). So-called congruency effects in behaviour are seen as a prototypical marker for the interaction between controlled and automatic processes. However, it is still unclear whether automatic and controlled processes interact over time with sensory and motor-specific processes. The present project aims to resolve this issue by investigating whether (a) automatic and controlled processes are integrated at an abstract, central stage of processing and as such should be independent from the specific sensory input (e.g., visual or audio) and motoric output (e.g., hand or foot) modalities, or (b) whether automatic and controlled processes interact at the level of sensory and/or motor processing. Starting from the premise of specific temporal-characteristics of automatic processes as proposed by the Diffusion Model for Conflict Tasks (DMC, Ulrich et al., 2015), we will combine sophisticated behavioral measures (e.g., distributional analyses) with experimental modality-specific manipulations (e.g., required motor force) to tackle the question concerning the integration locus or loci of activation generated by automatic and controlled processes as well as its potential task specificity.

Project leader: Dr. Victor Mittelstädt

Funding period: 06/2021-06/2024

The Experiential-Simulation View of Comprehension: How is Sentence Meaning composed?

(Project within the Collaborative Research Centres SFB 833: Emergence of Meaning: The Dynamics and Adaptivity of Linguistic Structures)

We focus on the ‘experiential simulations view of comprehension’. Our previous research showed that non-linguistic memory traces reflecting the comprehender’s experiences are automatically activated during word processing, and that comprehenders build simulations beyond the word level. No definite evidence was found for the assumption that simulations are functionally relevant for comprehension. In Phase III, we will continue investigating simulation processes at the sentence level, focusing particularly on the question regarding incrementality. We will also investigate the functional relevance of simulations for the processing of expressions requiring substantial composition processes, as well as for children’s language processing.

Project Leaders: Prof. Dr. Barbara Kaup, Prof. Dr. Hartmut Leuthold

Funding period: 2017-2021

 

MoLCINS: One- versus two-step models of language comprehension: Investigations employing negative sentences

(Project in DFG Priority Programme SPP 1727: XPrag.de: New Pragmatic Theories Based on Experimental Evidence)

There is an ongoing debate in psycholinguistics whether comprehension processes are best characterized by one- or by two-step models of comprehension. According to two-step models, comprehenders first compute the context-independent meaning of an expression based on their linguistic knowledge, and only in a second step take into account their general world knowledge as well as contextual and pragmatic information. In contrast, one-step models assume that various sources of linguistic and non-linguistic knowledge are taken into account simultaneously. In the present project, and on the basis of our findings from the first funding phase, we aim at contributing to this debate concerning the semantic/pragmatic interface by studying the processing of negative sentences. We will focus on the question whether aspects related to the pragmatics of negation are taken into account early on during comprehension, which is of central importance to the distinction between one- and two-step models of comprehension. In particular, we will present affirmative and negative sentences in the context of visual worlds that either do or do not license the use of negation in the sentences. We also aim to find out more about the exact conditions under which negative utterances are pragmatically felicitous. Finally, we are interested in comparing effects attributed to the pragmatics of negation with other pragmatic effects during comprehension, such as effects reflecting the integration of scalar implicatures. In order to gain insights into the temporal characteristics of the processes involved in comprehension, we plan to measure event-related brain potentials (ERPs) and eye-movements during reading and listening. We expect that this project will provide important insights into the comprehension processes at the semantic/pragmatic interface.

Project Leaders: Dr. Carolin Dudschig, Prof. Dr. Barbara Kaup, Prof. Dr. Hartmut Leuthold

Funding period: 2017-2020

The processing of moral transgressions: Investigating the role of affective evaluation

(DFG Project)

Currently, there is an ongoing debate concerning the functional role of emotion in moral judgment and decision-making (JDM). Thus, emotions may either (a) play a causal role, (b) exert a moderating influence, or (c) merely reflect an epiphenomenal by-product of moral JDM. Even if one assumes that emotions play a critical role in JDM, so far the mechanisms underlying their influence on moral information processing have remained unclear. In the present project, we are specifically interested in the question of whether affective processes and states assumed to be involved in moral JDM do indeed precede the decision process, as assumed by two prominent theories of moral judgment (social-intuitivist model of Haidt and dual-process theory of Greene and colleagues). Hence, we aim at examining the time-course of moral processing in paradigms that give us access to the underlying mental processes and (affective) states in close temporal proximity to their elicitation. A further goal is to assess the impact of emotion-related dimensions (arousal, valence, motivational tendency) on moral JDM for different types of moral (or socio-normative) materials. To address these questions, we employ different vignettes and task paradigms, analysing psychophysiological measures (event- related potentials, electrodermal activity, facial muscle activity) in addition to behavioural ones (response time).

Project Leader: Prof. Dr. Hartmut Leuthold

Funding period: 2016-2020