Mental State Reasoning Research Articles

Belief Files in Theory of Mind Reasoning

Humans seem to readily track their conspecifics’ mental states, such as their goals and beliefs from early infancy. However, the underlying cognitive architecture that enables such powerful abilities remains unclear. Here I will propose that a basic representational structure, the belief file, could provide the foundation for efficiently encoding, and updating information about, others’ beliefs in online social interactions. I will discuss the representational possibilities offered by the belief file and the ways in which the repertoire of mental state reasoning is shaped by the characteristics of its constituents. A series of questions will be outlined concerning the representational skeleton of the belief file, sketching a possible structure that supports the rapid encoding and re-identification of belief related information (e.g., variables for the agent, as the belief holder and for the belief-content). After analyzing the possible limitations of the belief attribution system, I will examine some of its characteristics that might enable a flexibility that is often neglected. I will suggest that operations involving belief files are not impeded by the absence of precise first-person information regarding their contents. In fact, the system permits manipulations with “empty” belief files, allowing humans to ascribe beliefs to conspecifics based on little or no direct information regarding the content of the mental state. Such an analysis aims to advance our understanding of how spontaneous belief attribution may be performed, and to provide an insight into the possible mechanisms that allow humans to successfully navigate the social world.

Anxious and egocentric: how specific emotions influence perspective taking.

People frequently feel anxious. Although prior research has extensively studied how feeling anxious shapes intrapsychic aspects of cognition, much less is known about how anxiety affects interpersonal aspects of cognition. Here, we examine the influence of incidental experiences of anxiety on perceptual and conceptual forms of perspective taking. Compared with participants experiencing other negative, high-arousal emotions (i.e., anger or disgust) or neutral feelings, anxious participants displayed greater egocentrism in their mental-state reasoning: They were more likely to describe an object using their own spatial perspective, had more difficulty resisting egocentric interference when identifying an object from others' spatial perspectives, and relied more heavily on privileged knowledge when inferring others' beliefs. Using both experimental-causal-chain and measurement-of-mediation approaches, we found that these effects were explained, in part, by uncertainty appraisal tendencies. Further supporting the role of uncertainty, a positive emotion associated with uncertainty (i.e., surprise) produced increases in egocentrism that were similar to anxiety. Collectively, the results suggest that incidentally experiencing emotions associated with uncertainty increase reliance on one's own egocentric perspective when reasoning about the mental states of others.

An individual task meta-analysis of social cognition in euthymic bipolar disorders

ObjectiveSocial cognition has been shown to be affected in bipolar disorders, even during euthymia. However, the social cognitive profile of this group of disorders remains to be ascertained, given that such a broad neuropsychological construct has not been systematically examined in bipolar subjects across different tasks. The aim of this study was to quantify the magnitude of patient-control differences for distinct social cognition assessment instruments: the Hinting Task, the Eyes Test, Faux Pas, the Mayer–Salovey–Caruso Emotional Intelligence Test, and emotional labeling using visual stimuli. MethodEffect sizes were extracted from studies chosen according to more stringent criteria than previously used in systematic reviews on the topic and pooled by means of meta-analytical procedures. ResultsNo significant patient-control differences were found for the recognition of three basic emotions (happiness, sadness, and anger). Small but significant effect sizes favoring healthy controls (Hedges׳ g<0.5) were noted for emotional intelligence, the Hinting Task, the Eyes Test, and the recognition of fear, disgust, and surprise. A medium effect size (Hedges׳ g=0.58) was noted for the Faux Pas Test. LimitationsThe possible effects of other neurocognitive impairments on social cognitive performance could not be explored. ConclusionOn average, small-to-moderate differences may exist between euthymic bipolar disorder subjects and healthy controls regarding social cognitive performance, with mental state decoding being more preserved than mental state reasoning. The influence of clinical and neurocognitive variables, which may play an important role in the social cognitive outcomes of these patients, deserves further clarification.

Gender and Theory of Mind in Preschoolers’ Group Effort: Evidence for Timing Differences Behind Children’s Earliest Social Loafing

This study explored mental state reasoning within the context of group effort and possible differences in development between boys and girls. Preschool children (59 girls, 47 boys) were assessed for theory of mind (ToM) ability using classic false belief tests. Children participated in group effort conditions that alternated from one condition, where individual effort was transparent and obvious, to one where individual effort remained anonymous. The aim was to investigate if emergent mental state reasoning, after controlling for age, was associated with the well-known phenomenon of reduced effort in group tasks (“social loafing”). Girls had slightly higher ToM scores and social loafing than boys. Hierarchical regression, controlling for age, indicated that understanding of others’ false beliefs uniquely predicted social loafing and interacted weakly with gender status.

Are all beliefs equal? Implicit belief attributions recruiting core brain regions of theory of mind.

Humans possess efficient mechanisms to behave adaptively in social contexts. They ascribe goals and beliefs to others and use these for behavioural predictions. Researchers argued for two separate mental attribution systems: an implicit and automatic one involved in online interactions, and an explicit one mainly used in offline deliberations. However, the underlying mechanisms of these systems and the types of beliefs represented in the implicit system are still unclear. Using neuroimaging methods, we show that the right temporo-parietal junction and the medial prefrontal cortex, brain regions consistently found to be involved in explicit mental state reasoning, are also recruited by spontaneous belief tracking. While the medial prefrontal cortex was more active when both the participant and another agent believed an object to be at a specific location, the right temporo-parietal junction was selectively activated during tracking the false beliefs of another agent about the presence, but not the absence of objects. While humans can explicitly attribute to a conspecific any possible belief they themselves can entertain, implicit belief tracking seems to be restricted to beliefs with specific contents, a content selectivity that may reflect a crucial functional characteristic and signature property of implicit belief attribution.

Capacities for theory of mind, metacognition, and neurocognitive function are independently related to emotional recognition in schizophrenia

While many with schizophrenia spectrum disorders experience difficulties understanding the feelings of others, little is known about the psychological antecedents of these deficits. To explore these issues we examined whether deficits in mental state decoding, mental state reasoning and metacognitive capacity predict performance on an emotion recognition task. Participants were 115 adults with a schizophrenia spectrum disorder and 58 adults with substance use disorders but no history of a diagnosis of psychosis who completed the Eyes and Hinting Test. Metacognitive capacity was assessed using the Metacognitive Assessment Scale Abbreviated and emotion recognition was assessed using the Bell Lysaker Emotion Recognition Test. Results revealed that the schizophrenia patients performed more poorly than controls on tests of emotion recognition, mental state decoding, mental state reasoning and metacognition. Lesser capacities for mental state decoding, mental state reasoning and metacognition were all uniquely related emotion recognition within the schizophrenia group even after controlling for neurocognition and symptoms in a stepwise multiple regression. Results suggest that deficits in emotion recognition in schizophrenia may partly result from a combination of impairments in the ability to judge the cognitive and affective states of others and difficulties forming complex representations of self and others.

The Good, the Bad, and the Just: Justice Sensitivity Predicts Neural Response during Moral Evaluation of Actions Performed by Others

Morality is a fundamental component of human cultures and has been defined as prescriptive norms regarding how people should treat one another, including concepts such as justice, fairness, and rights. Using fMRI, the current study examined the extent to which dispositions in justice sensitivity (i.e., how individuals react to experiences of injustice and unfairness) predict behavioral ratings of praise and blame and how they modulate the online neural response and functional connectivity when participants evaluate morally laden (good and bad) everyday actions. Justice sensitivity did not impact the neuro-hemodynamic response in the action-observation network but instead influenced higher-order computational nodes in the right temporoparietal junction (rTPJ), right dorsolateral and dorsomedial prefrontal cortex (rdlPFC, dmPFC) that process mental states understanding and maintain goal representations. Activity in these regions predicted praise and blame ratings. Further, the hemodynamic response in rTPJ showed a differentiation between good and bad actions 2 s before the response in rdlPFC. Evaluation of good actions was specifically associated with enhanced activity in dorsal striatum and increased the functional coupling between the rTPJ and the anterior cingulate cortex. Together, this study provides important knowledge in how individual differences in justice sensitivity impact neural computations that support psychological processes involved in moral judgment and mental-state reasoning.

A pilot study: Investigating the neural correlates of social cognition in neuropsychiatric disorders using fMRI

Event Abstract Back to Event A pilot study: Investigating the neural correlates of social cognition in neuropsychiatric disorders using fMRI Charlotte Knee-Zaska1*, Omar Mothersill1 and Gary Donohoe1 1 NUI Galway, Psychology, Ireland Social cognition can be defined as mental processes allowing people to perceive and process information about themselves and others, enabling successful navigation of the social world (Van Overwalle, 2009). Social cognitive defects (such as emotion recognition and mental state reasoning, or ‘theory of mind’) are a core feature of several psychiatric disorders, including schizophrenia (Couture et al., 2006) and strongly predict functional outcome (Fett et al., 2011). For this reason social cognition is a potential target for new pharmacological and/or cognitive treatments. A better understanding of the neurobiological mechanisms involved are a critcal step towards this. The present study aims to elucidate these mechanisms using functional MRI across a sample of psychosis patients (N = 30) and demographically-matched healthy controls (N = 30). Three robust social cognitive tasks will be used during fMRI, including (1) emotion recognition, where participants observe video clips of angry or neutral facial expressions (Grosbras and Paus, 2006), (2) theory of mind, where participants view cartoons and contemplate the mental states of the characters (Brüne et al., 2008) and (3) social exclusion, where participants experience social exclusion from a computer ball game (Sebastian et al., 2011).This study will also examine effects of putative risk factors for psychosis, including effects of genome-wide associated genetic risk variants (Ripke et al., 2011) and environmental factors. Using a single healthy volunteer, this pilot fMRI study highlighted several brain regions that were activated during each of these social cognitive tasks at a p<0.05 level, including corticolimbic and default mode networks, consistent with previous social neuroimaging studies. References Van Overwalle, F. (2009). Social cognition and the brain: a meta‐analysis.Human brain mapping, 30(3), 829-858. Couture, S. M., Penn, D. L., & Roberts, D. L. (2006). The functional significance of social cognition in schizophrenia: a review. Schizophrenia bulletin, 32(suppl 1), S44-S63. Fett, A. K. J., Viechtbauer, W., Dominguez, M. D. G., Penn, D. L., van Os, J., & Krabbendam, L. (2011). The relationship between neurocognition and social cognition with functional outcomes in schizophrenia: a meta-analysis. Neuroscience & Biobehavioral Reviews, 35(3), 573-588. Grosbras, M. H., & Paus, T. (2006). Brain networks involved in viewing angry hands or faces. Cerebral Cortex, 16(8), 1087-1096. Brüne, M., Lissek, S., Fuchs, N., Witthaus, H., Peters, S., Nicolas, V., ... & Tegenthoff, M. (2008). An fMRI study of theory of mind in schizophrenic patients with “passivity” symptoms. Neuropsychologia, 46(7), 1992-2001. Sebastian, C. L., Tan, G. C., Roiser, J. P., Viding, E., Dumontheil, I., & Blakemore, S. J. (2011). Developmental influences on the neural bases of responses to social rejection: implications of social neuroscience for education. Neuroimage, 57(3), 686-694. Keywords: social cognition, functional magnetic resonance imaging (fMRI), Schizophrenia, theory of mind (ToM), Emotion recogntion, social exclusion Conference: Neuroscience Ireland Young Neuroscientists Symposium 2014 , Dublin, Ireland, 20 Sep - 20 Sep, 2014. Presentation Type: Poster Presentation Topic: Early Career Neuroscience Citation: Knee-Zaska C, Mothersill O and Donohoe G (2014). A pilot study: Investigating the neural correlates of social cognition in neuropsychiatric disorders using fMRI . Front. Neurosci. Conference Abstract: Neuroscience Ireland Young Neuroscientists Symposium 2014 . doi: 10.3389/conf.fnins.2014.87.00020 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 12 Sep 2014; Published Online: 12 Sep 2014. * Correspondence: Mrs. Charlotte Knee-Zaska, NUI Galway, Psychology, Galway, Ireland, kneec@tcd.ie Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract The Authors in Frontiers Charlotte Knee-Zaska Omar Mothersill Gary Donohoe Google Charlotte Knee-Zaska Omar Mothersill Gary Donohoe Google Scholar Charlotte Knee-Zaska Omar Mothersill Gary Donohoe PubMed Charlotte Knee-Zaska Omar Mothersill Gary Donohoe Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.

Using Fiction to Assess Mental State Understanding: A New Task for Assessing Theory of Mind in Adults

Social functioning depends on the ability to attribute and reason about the mental states of others – an ability known as theory of mind (ToM). Research in this field is limited by the use of tasks in which ceiling effects are ubiquitous, rendering them insensitive to individual differences in ToM ability and instances of subtle ToM impairment. Here, we present data from a new ToM task – the Short Story Task (SST) - intended to improve upon many aspects of existing ToM measures. More specifically, the SST was designed to: (a) assess the full range of individual differences in ToM ability without suffering from ceiling effects; (b) incorporate a range of mental states of differing complexity, including epistemic states, affective states, and intentions to be inferred from a first- and second-order level; (c) use ToM stimuli representative of real-world social interactions; (d) require participants to utilize social context when making mental state inferences; (e) exhibit adequate psychometric properties; and (f) be quick and easy to administer and score. In the task, participants read a short story and were asked questions that assessed explicit mental state reasoning, spontaneous mental state inference, and comprehension of the non-mental aspects of the story. Responses were scored according to a rubric that assigned greater points for accurate mental state attributions that included multiple characters’ mental states. Results demonstrate that the SST is sensitive to variation in ToM ability, can be accurately scored by multiple raters, and exhibits concurrent validity with other social cognitive tasks. The results support the effectiveness of this new measure of ToM in the study of social cognition. The findings are also consistent with studies demonstrating significant relationships among narrative transportation, ToM, and the reading of fiction. Together, the data indicate that reading fiction may be an avenue for improving ToM ability.

When Mental States Matter, When They Don't, and What That Means for Morality

AbstractResearch has shown that moral judgments depend on the capacity to engage in mental state reasoning. In this article, we will first review behavioral and neural evidence for the role of mental states (e.g., people's beliefs, desires, intentions) in judgments of right and wrong. Second, we will consider cases where mental states appear at first to matter less (i.e., when people assign moral blame for accidents and when explicit information about mental states is missing). Third, we will consider cases where mental states, in fact, matter less, specifically, in cases of “purity” violations (e.g., committing incest, consuming taboo foods). We will discuss how and why mental states do not matter equivalently across the multi‐dimensional space of morality. In the fourth section of this article, we will elaborate on the possibility that norms against harmful actions and norms against “impure” actions serve distinct functions – for regulating interpersonal interactions (i.e., harm) versus for protecting the self (i.e., purity). In the fifth and final section, we will speculate on possible differences in how we represent and reason about other people's mental states versus our own beliefs and intentions. In addressing these issues, we aim to provide insight into the complex structure and distinct functions of mental state reasoning and moral cognition. We conclude that mental state reasoning allows us to make sense of other moral agents in order to understand their past actions, to predict their future behavior, and to evaluate them as potential friends or foes.

Malice aforethought?

Distinguishing between intentional and unintentional harm relies on mental state reasoning, a faculty that is impaired in high-functioning autistic people. The right temporoparietal junction (RTPJ) is highly active when making moral judgements. Multivoxel pattern analysis (MVPA) revealed that variations in responses in RTPJ across voxels correlated with the moral judgement taken by an individual; these patterns were absent in autistic subjects. The data suggest that MVPA can be used to gain insight into how the RTPJ represents features of beliefs that are relevant for moral decision-making and understanding underlying intent.

Decoding moral judgments from neural representations of intentions

Intentional harms are typically judged to be morally worse than accidental harms. Distinguishing between intentional harms and accidents depends on the capacity for mental state reasoning (i.e., reasoning about beliefs and intentions), which is supported by a group of brain regions including the right temporo-parietal junction (RTPJ). Prior research has found that interfering with activity in RTPJ can impair mental state reasoning for moral judgment and that high-functioning individuals with autism spectrum disorders make moral judgments based less on intent information than neurotypical participants. Three experiments, using multivoxel pattern analysis, find that (i) in neurotypical adults, the RTPJ shows reliable and distinct spatial patterns of responses across voxels for intentional vs. accidental harms, and (ii) individual differences in this neural pattern predict differences in participants' moral judgments. These effects are specific to RTPJ. By contrast, (iii) this distinction was absent in adults with autism spectrum disorders. We conclude that multivoxel pattern analysis can detect features of mental state representations (e.g., intent), and that the corresponding neural patterns are behaviorally and clinically relevant.

Rethinking theory of mind in high‐functioning autism spectrum disorder

The sociocommunicative problems in autism spectrum disorder (ASD) are traditionally linked to impairments in theory of mind (ToM), the ability to ascribe mental states to others. Although ToM impairments are consistently reported in young children with ASD, findings on more advanced ToM understanding in older individuals with high-functioning ASD (HFASD) are less straightforward. Therefore, we assessed the advanced ToM abilities of a large sample of school-aged children and adolescents with HFASD (n = 194; 6-20 years) and compared them to a typically developing (TD) comparison group (n = 60). Participants' advanced ToM was assessed with five social stories containing second-order false beliefs, display rules, double bluff, faux pas, and sarcasm. Participants with HFASD performed equally well on each of the ToM stories as their TD peers. Consistent age effects were noticed with adolescents outperforming the children. Furthermore, advanced ToM was positively associated with participants' age, verbal abilities, and general reasoning abilities. Counter to what the ToM theory of ASD would predict, school-aged children and adolescents with HFASD seem to be able to master the theoretical principles of advanced mental state reasoning. However, they may still fail to apply these theoretical principles during everyday social interactions.

Social cognition in bipolar disorder versus schizophrenia: comparability in mental state decoding deficits

Neuropsychological studies comparing patients with bipolar disorder (BD) to patients with schizophrenia (SZ) suggest milder cognitive deficits in BD patients and across a smaller range of functions. The present study investigated whether this pattern is also true for social cognition - a range of socially relevant abilities, including emotion perception and recognition, theory of mind, and social attributions - by comparing performance on measures of social cognition in patients with BD, SZ, and healthy participants. One hundred and two patients with BD, 208 patients with SZ, and 132 healthy participants were assessed using a battery of tasks measuring basic neuropsychological and social cognition. We observed significant differences between patients with BD and healthy participants in a test of mental state decoding ('eyes task') that was at a level comparable to deficits seen in patients with SZ. By comparison, BD patients showed more subtle deficits in mental state reasoning ('hinting task') than those shown by patients with SZ. Mental state decoding difficulties are significant in BD. An important direction for further research will be to establish to what extent these deficits affect social and occupational functioning as a potential target for therapeutic intervention.

Towards an integrative approach to understanding quality of life in schizophrenia: the role of neurocognition, social cognition, and psychopathology

The term “schizophrenia” refers to a debilitating group of disorders that usually results in a severely impaired quality of life (QoL). Symptomatology appears to have a substantial role in determining QoL, although the relationship between QoL and specific psychotic symptoms is still unclear and has demonstrated mixed results. Due to the intrinsic importance of social functioning in QoL, and the mediating effect of social cognition on social functioning, the aim of this study was to try to investigate QoL in schizophrenia, not only in terms of symptomatology, but also in consideration of potential neurocognitive and social cognitive contributing factors. MethodsTwenty-eight clinically stable patients with schizophrenia performed a broad range of neurocognitive and social cognitive assessments, and also participated in a semi-structured interview of QoL, assessing four partially independent subdomains of QoL. A stepwise regression model was used to determine the best predictors of QoL, and additionally a mediator analysis was performed to test for the mediating power of social cognition on QoL. ResultsNegative symptoms, intelligence, executive functioning and social cognition all had some power in predicting QoL in schizophrenia. Though most interestingly, mental state reasoning was specifically found to be most strongly related with the Intrapsychic Foundation subdomain of QoL, whereas neurocognition and symptom severity were associated with other subdomains of QoL. ConclusionsThe association between mental state reasoning and the more “internal” aspects of QoL in schizophrenia may reflect a specific role for social cognition in introspective and subjective judgments of one's own QoL, whereas neurocognition and negative symptomatology may be more predictive of the external or extrinsic aspects of QoL. In conclusion, social cognitive skills appear to play a crucial role in the experience of one's own subjective well-being, which could help to explain previous inconsistencies in the literature investigating QoL in schizophrenia.

Electrophysiological difference between mental state decoding and mental state reasoning

Previous studies have explored the neural mechanism of Theory of Mind (ToM), but the neural correlates of its two components, mental state decoding and mental state reasoning, remain unclear. In the present study, participants were presented with various photographs, showing an actor looking at 1 of 2 objects, either with a happy or an unhappy expression. They were asked to either decode the emotion of the actor (mental state decoding task), predict which object would be chosen by the actor (mental state reasoning task), or judge at which object the actor was gazing (physical task), while scalp potentials were recorded. Results showed that (1) the reasoning task elicited an earlier N2 peak than the decoding task did over the prefrontal scalp sites; and (2) during the late positive component (240–440ms), the reasoning task elicited a more positive deflection than the other two tasks did at the prefrontal scalp sites. In addition, neither the decoding task nor the reasoning task has no left/right hemisphere difference. These findings imply that mental state reasoning differs from mental state decoding early (210ms) after stimulus onset, and that the prefrontal lobe is the neural basis of mental state reasoning.

Why Studies of Autism Spectrum Disorders Have Failed to Resolve the Theory Theory Versus Simulation Theory Debate

The Theory Theory (TT) versus Simulation Theory (ST) debate is primarily concerned with how we understand others’ mental states. Theory theorists claim we do this using rules that are akin to theoretical laws, whereas simulation theorists claim we use our own minds to imagine ourselves in another’s position. Theorists from both camps suggest a consideration of individuals with autism spectrum disorders (ASD) can help resolve the TT/ST debate (e.g., Baron-Cohen 1995; Carruthers 1996a; Goldman 2006). We present a three-part argument that such research has so far been inconclusive and that the prospects for studies of ASD to resolve the debate in the near future remain uncertain. First, we discuss evidence indicating that some individuals with ASD can perform effectively on tests of mental state understanding, which questions what ASD can tell us regarding theorising or simulation. Second, we claim that there is compelling evidence that domain-general mechanisms are implicated in mental state reasoning, which undermines how ASD might inform the TT/ST debate given that both theories appeal to domain-specific mindreading mechanisms. Third, we suggest that neuroscientific evidence for an assumed role of the mirror neuron system in autism also fails to arbitrate between TT and ST. We suggest that while the study of ASD may eventually provide a resolution to the TT/ST debate, it is also vital for researchers to examine the issues through other avenues, for example, by examining people’s everyday counterfactual reasoning with mental state scenarios.

Moral Universals and Individual Differences

Contemporary moral psychology has focused on the notion of a universal moral sense, robust to individual and cultural differences. Yet recent evidence has revealed individual differences in the psychological processes for moral judgment: controlled cognition, mental-state reasoning, and emotional responding. We discuss this evidence and its relation to cross-cultural diversity in morality.

Do deaf adults with limited language have advanced theory of mind?

Previous studies show that deaf children have deficits in false belief understanding due to their language impairment. However, it is not clear whether deaf adults still have problems in advanced theory of mind (ToM). The present study examined deaf adults’ performance on three aspects of advanced ToM. All of the deaf groups lacking mental state language tended to perform worse than the hearing group on explicit mental state understanding. Deaf groups with either vocabulary skill or interpersonal experience from early years were similar to the hearing group in implicit mental state reasoning. Individuals frequently using syntactic complements or having interpersonal experience with hearing people from early years tended to use ToM better. Moreover, language ability was the only predictor for explicit rather than implicit mental state understanding. Sufficient language is not necessary for all aspects of advanced ToM. Rich interpersonal experience as a substitute for language may facilitate deaf adults’ advanced ToM.

What gets the attention of the temporo-parietal junction? An fMRI investigation of attention and theory of mind

Functional magnetic resonance imaging (fMRI) studies have demonstrated a critical role for a cortical region in the right temporo-parietal junction (RTPJ) in “theory of mind” (ToM), or mental state reasoning. In other research, the RTPJ has been implicated in the deployment of attention to an unexpected stimulus. One hypothesis (“attention hypothesis”) is that patterns of RTPJ activation in ToM tasks can be fully explained by appeal to attention: stimuli that apparently manipulate aspects of ToM are in fact manipulating aspects of attention. On an alternative hypothesis (“ToM hypothesis”), functional regions identified by ToM tasks are selective for ToM, and not just for any unexpected stimulus. Here, we used fMRI to test these competing hypotheses: are brain regions implicated in ToM, including the RTPJ, LTPJ, and precuneus, recruited specifically for mental states, or for any unexpected stimulus? We first identified brain regions implicated in ToM, using a standard paradigm: participants read stories about false beliefs and false physical representations (e.g., outdated photographs). Participants also read a new set of stories describing mental or physical states, which were unexpected or expected. Regions of interest analyses revealed a higher response in the RTPJ, LTPJ, and precuneus, for mental versus physical stories, but no difference for unexpected and expected stories. Whole-brain random effects analyses also revealed higher activation in these regions for mental versus physical stories. This pattern provides evidence for the ToM hypothesis: the response in these functional regions is selective for mental state content, whether that content is unexpected or expected.