Jan Frederik Sima
|Title||Dr. rer. nat.|
I was a doctoral researcher in the project R1-[ImageSpace]. I'm interested in theories, architectures, and (computational) models of cognition. In particular, I investigated the mental processes underlying mental imagery. Related to this topic, I also dealt with the relationship between eye movements and mental processes during spatial reasoning.
Abstract of my Ph.D. thesis:
The thesis develops a new theory of visuo-spatial mental imagery. The theory is concretized in a formal framework and implemented as a computational model. The theory and its model are evaluated against a set of empirical phenomena and compared to the contemporary theories of mental imagery. The new theory is shown to provide explanations for the considered phenomena that partly go beyond those of the contemporary theories.
The thesis is motivated by two main observations.
First, the observation that the lack of formalization of the current psychological and philosophical theories of mental imagery limits the progress of the imagery debate, i.e., the question about the nature of mental imagery. A formalized theory is able to provide more detailed explanations and predictions for the empirical data which can facilitate further empirical studies. Furthermore, sufficiently formalized theories become comparable with objective measures thus making similarities and differences between theories more transparent.
Second, some of the contemporary theories of mental imagery stress the involvement of rich mental representations in cognition and mental imagery. This approach has been considered problematic with respect to more recent results such as the functionality of eye movements during mental imagery as well as the neuropsychological findings on unilateral neglect. The enactive theory poses an exception and stresses the importance of sensorimotor interactions for mental imagery.
The new theory shares assumptions with the enactive theory with respect to direct and active vision and the relationship between vision and imagery. It combines this view with grounded mental concepts which function as hubs to low-level perceptional actions. The theory understands the process of mental imagery in the context of internal simulations of sensorimotor interactions. Mental images are based on grounded concepts whose semantics are made explicit by the overt and covert employment of the low-level perceptual actions they link to. This employment of perceptual actions makes low-level perceptual information available which represents an instance of the conceptually described mental image. Critically, this perceptual information is not made available by an activation of early visual areas but by mechanisms of proprioception and anticipation.
You can find the pdf here
And a html version here
If you have remarks, corrections, additions, etc. regarding the thesis, send me an email: jfsima A-T gmail D-O-T com.
DFG SFB/TR-8 Spatial Cognition: Project R1-[ImageSpace] - Mental Representations of Spatial Environments