Kognitive und kortikale Korrelate und computationale Modellierung der Überzeugungsrevision beim räumlichen Schließen
Zusammenfassung der Projektergebnisse
People constantly change their beliefs when confronted with new contradicting information. Especially during these days, while the Covid-19 disease spreads around the globe, it has become obvious how important it is to take contradicting pieces of information into account and to understand the basic processes of belief revision. But, one domain of belief revision has been widely neglected for a long time: the spatial. Therefore, the first funding period of this project introduced spatial belief revision as a new research topic. We investigated the basic cognitive mechanisms underlying these revision choices. Two major novel findings arose from the second funding period: First, against previous assumptions people do not construct a completely new model from scratch when confronted with an incontrovertible fact. They rather revise their beliefs by varying an initially constructed preferred mental model (PMM). Thus, the revision is driven by model variation and is not equally based on all three phases of spatial reasoning (construction, inspection, variation, see report of the first funding period for detailed information). The second finding was even more surprising even for us. People do not base their belief revision choices on the principle of epistemic entrenchment, which is well known from the theories of rule-based reasoning, nor solely on the principle of minimal change, which is the common rationale for belief revision derived from AI research. In fact, the model variation – and therefore the belief revision – is largely driven by the linguistic asymmetries of the facts` objects. Consider a person who constructed a spatial belief of the relation of three objects: Motorcycle Car Bike. When confronted with an indisputable fact like “The Bike is to the left of the Motorcycle” there are two logically equivalent belief revision options. You can either move the bike to the left side of the arrangement or the Motorcycle to the right: Bike Motorcycle Car or Car Bike Motorcycle. Following the logical equivalence of both options, people should choose one of both arrangements simply by chance. However, we could show that there is a clear preference. In more than 90% of the cases, people choose the first arrangement (Bike Motorcycle Car). Looking at the fact (The Bike is to the left of the Motorcycle), the first mentioned object (bike) is seen as more flexible. Therefore, the bike is located in relation to the other object (Motorcycle). In this example, the bike is the “to be located object” (LO) and the house the “reference object” (RO). This linguistic asymmetry of objects is well known in spatial language. But we were the first who identified a direct influence of this asymmetry on the cognitive processes of spatial belief revision. We call this very robust effect the LO-principle. In this second funding period we tested both, the implications and the boundaries of this LO-principle. We extended our basic paradigm and took a broader view on spatial reasoning and belief revision. We tested the role of embodied cognition, language, presentation format, multimodality and social aspects for revision choices. In five work packages, we showed that the basic spatial belief mechanisms like the LO-principle have cross-cultural validity, and can be influenced by background knowledge about object size or mobility. We were successful in showing that the theories of mental models and embodied cognition can supplement each other well. Furthermore, we integrated the findings of the first and second funding period in a comprehensive cognitive theory and computational model about spatial reasoning and belief revision. This theory and findings were underpinned with experiments using functional magnetic resonance imaging (fMRI) and transcranial brain stimulation (TMS) and we identified the brain areas relevant for spatial reasoning and belief revision. Many results of the project have been incorporated into an MIT Press book by Knauff, in which a comprehensive theory of human spatial thinking is developed. Our experiments with individuals served as the ideal test bed for the basic principles of spatial belief construction and revision processes. However, belief revision does not solely take place in one mind. Rather, these processes can be regarded as highly social processes, embedded in discussions and argumentation. Thanks to the long project duration, we had the time to investigate these social aspects of spatial reasoning and belief revision in addition to the planned working program of the project. We tested, whether groups of two persons (dyads) can construct and (if necessary) vary a spatial mental model collaboratively. On the one hand, dyads were able to outperform individuals' reasoning capabilities. On the other hand, groups could produce a conformity pressure on revision choices and diminish the LO-principle. Overall, these findings complement our findings very well with a more realistically social setting. Altogether, we were very successful in creating a whole new paradigm for experimental research of spatial reasoning, leading to many new insights about this complex and yet basic principles of human thinking.
Projektbezogene Publikationen (Auswahl)
-
(2013). A theory and a computational model of spatial reasoning with preferred models, Psychological Review, 120, 561-588
Ragni, M., & Knauff, M.
-
(2013). Spatial belief revision. Journal of Cognitive Psychology, 25, 147-156
Knauff, M., Bucher, L., Krumnack, A., & Nejašmić, J.
-
(2014). Belief revision and way-finding. Cognitive Processing, 15(1), 99-106
Bucher, L., Röser, F., Nejašmić, J., & Hamburger, K.
-
(2015). Grounded spatial belief revision. Acta Psychologica, 157, 144-154
Nejašmić, J., Bucher, L., & Knauff, M.
-
(2015). The construction of spatial mental models – A new view on the continuity effect. The Quarterly Journal of Experimental Psychology, 68(9), 1794-1812
Nejašmić, J., Bucher, L., & Knauff, M.
-
(2016). Uncertain relational reasoning in the parietal cortex. Brain and Cognition, 104, 72-81
Ragni, M., Franzmeier, I., Maier, S., & Knauff, M.
-
(2018). TMS applied to V1 can facilitate reasoning. Experimental Brain Research, 236(8), 2277-2286
Hamburger, K., Ragni, M., Karimpur, H., Franzmeier, I., Wedell, F., & Knauff, M.
