Matching mind to world and vice versa: Functional dissociations between belief and desire mental state processing

Abstract

With the aim of understanding how different mental or intentional states are processed in the brain, the present functional magnetic resonance imaging (fMRI) study examined the brain correlates during the ascription of belief intentional states relative to desire intentional states as well as the effect of incongruent relative to congruent intentional states. To this end, sentences containing scenarios were presented to participants and their task was to make judgments concerning the ascription of intentional states based on this information. Belief ascriptions, relative to desire ascriptions, were accompanied by activations in lateral prefrontal structures that include areas known to be involved in relational and conceptual reasoning. Desire ascriptions, in contrast, were accompanied by activations in regions of the medial prefrontal cortex, superior temporal gyri and hippocampal formation, all of which are known for their involvement in self-referential, autobiographical and episodic memory-relevant processes. In addition, the ascription of intentional states that were incongruent with reality (false beliefs and unfulfilled desires) was compared to the ascription of intentional states that were congruent to reality (true belief and fulfilled desires). While no brain region was selectively activated during the processing of unfulfilled desires, the processing of false beliefs was associated with stronger activations in the dorsal medial prefrontal cortex, an area that has been previously linked to the process of decoupling in false belief attribution. These findings provide new insights into more fine-grained aspects of mental state reasoning.

Keywords:

• Theory of mind
• Mentalizing
• Functional neuroimaging
• Folk psychology
• Belief–desire reasoning

Acknowledgements

The authors thank Sylvia Mössinger for her assistance during the pilot studies and Heike Schmidt-Duderstedt for her assistance with the figures. This study was funded by the Max Planck Society (MPG).

Notes

¹The only mention of an experimental contrast between beliefs and desires within the same paradigm was in a review article by Saxe et al. (2004a, p. 109), where it was reported that the TPJ was significantly more responsive to beliefs than to desires in one previous study (Saxe & Kanwisher, 2003). However, it was also stated within the same context that the stimuli used in this study and others (that investigated the attribution of desires, goals and intentions) were suboptimal because “none of the stimuli in these studies were designed to exclude belief attribution.” Saxe (2006) has argued for the role played by the TPJ in underlying the representational aspects of mental states such as beliefs, and has also provided support for a special role played by the right TPJ in desire processing (Saxe & Wexler, 2005).

²We introduce the term “intentional states” here to avoid misunderstandings that could stem from the more generalized usage of the term “mental states” in social neuroscience, which is used to refer a wide variety of phenomena under the general umbrella of mentalizing or mental state reasoning. We employ the term “intentional states” to refer to specific mental states such as beliefs and desires that are held, within the purview of philosophy of mind, to characteristically exhibit “intentionality” or the capacity of the mind to be about or to represent things, properties and states of affairs (Dennett & Haugeland, 1987; Perry, 1994).

³As this was an unexpected finding and only approaching significance, we do not discuss this result further.

⁴Events pertaining to the scenario stage (where mental state information is being conveyed) and the question stage (where explicit mental state reasoning on part of the participant is required) were analyzed separately because the similarities and differences between the recruitment of brain regions as a function of which stage of mental state reasoning one is engaged in have not been systematically explored so far.

⁵There were also additional activations during desire relative to belief reasoning in occipital and fusiform regions. These unexpected results could reflect condition specific lower-level phenomena, such as different eye movement trajectories as a function of different reading strategies, or higher-level phenomena such as greater visual imagery. Further experiments and behavioral indices will be required to verify which of these alternatives is more plausible.