What is theory of mind and how does it affect a child if they have it or not?

Since the early 1980s there has been a surge of interest in the cognitive basis of our common sense ‘theory of mind’ and how it develops. According to our common sense, other people act because they have mental states of various kinds, for example, intentions, desires, beliefs, hopes, etc.; furthermore, such states have contents, for example, the belief that it is raining has the content ‘it is raining’ and the desire to avoid paying taxes has the content ‘avoid paying taxes.’ Contents individuate mental states (for example, determine which belief or which desire a person has) and play a critical role in causing behavior. Even preschool children have been found to attribute these sorts of mental states to other people and to treat these states-with-contents as causes of behavior. Children who are learning disabled can also employ such a theory of mind, suggesting that it is not the result of a general intellectual ability. Children with autistic disorder are selectively impaired in theory of mind ability even though their general intellectual abilities may be normal. Theory of mind is a key cognitive ability comprising the human social instinct. The psychological mechanisms underlying this ability are of intense current interest but their nature remains controversial.

ToM assessment measures for ASD disorders

ToM task batteries are important because they indicate that there is more to ToM than false-belief understanding and because they have the potential to highlight the specific ToM strengths and challenges that an individual brings to the social problem-solving situation. As on the false-belief task, performance on more comprehensive batteries may be influenced by attention, memory, linguistic, motivational, and situational factors (Tager-Flusberg, 2000). The impact of these shortcomings varies with the individual and the assessment procedures employed. To avoid these shortcomings, Hutchins et al. (2008) developed a psychometrically sound informant measure, Perceptions of Children’s Theory of Mind Measure (PCToMM). The PCToMM-E was developed to reflect variation in the theoretical background and assessment procedures in ToM research with typically developing children and children with ASD. The measure was designed to serve as an index of caregivers’ perceptions of children’s ToM knowledge, as well as children’s actual ToM knowledge. The developers characterize summated and averaged scores as yielding interval data that reflect a general composite of a child’s ToM knowledge based on more specific component variables. The PCToMM-E consists of 33 statements. The content of the PCToMM-E was guided by a review of the ToM literature, and items were developed to reflect the diverse theoretical perspectives. Scores demonstrated high test-retest reliability and correlated with verbal mental age and ToM task battery performance. Hutchins et al. (2008) indicate that when considering evaluation of individuals with ASD, there is a need to complement a selection of more traditional tasks with qualitative and observational data.

Hutchins and Prelock (2008), in their efforts to support the development of ToM of individuals with ASD, have found a variety of qualitative assessment methods. These include: (1) observation during naturalistic activities and routines and in more formal and structured testing situations and (2) triangulation to seek the impressions of parents, educators, and other professionals who know the individual well.

Mentalizing and ToM

ToM involves the ability to attribute mental states to oneself and others (e.g., understanding others’ actions, feelings, and intentions), and has been central to the schizotypy construct since its early formation (Langdon & Coltheart, 1999). Links between performance on ToM tasks, often developed for individuals with autism spectrum pathology, and schizotypal traits have provided evidence that only some aspects of ToM are affected in schizotypy. For example, schizotypy scores (reflecting positive, negative, and disorganization traits) have been associated with poor performance on a ToM task requiring recognition of social faux pas (Morrison, Brown, & Cohen, 2013), intact performance on measures of sarcasm (Jahshan & Sergi, 2007), and enhanced performance on a ToM task that involved detection of irony (McCleery et al., 2012). It is clear that behaviorally, high and low schizotypy cannot be reliably differentiated from each other based solely on ToM. However, there is substantial and consistent evidence for deficits in ToM in schizophrenia, in both populations in remission and those who are not, regardless of task (Bora, Yucel, & Pantelis, 2009). ToM seems to be an area where populations with schizophrenia are more severely impacted than populations with milder forms of schizotypy, who in some cases, may outperform controls.

Theory of Mind

ToM is the ability to infer and understand another's mental state (the beliefs, thoughts, intentions and feelings of another), and use this information to explain and predict human behavior. ToM is thought to be mediated mainly by the DMN. The vmPFC, precuneus and TPJ are considered to represent core ToM regions, and have been found to be activated across a wide range of different ToM tasks (Atique et al., 2011). There is some evidence that distinct subregions of these structures underlie affective ToM (i.e., “I understand what you feel”) versus cognitive ToM (i.e., “I understand your intentions”). For example, the TPJ and vmPFC are thought to be more relevant for cognitive ToM, while the precuneus and the dorsal mPFC may be more important for affective ToM (Atique et al., 2011; Frith and Frith, 2006).

The PCC, pSTS and temporal poles have been associated with particular aspects of ToM. For example the pSTS, a region thought to be important for biological motion processing, appears to be primarily activated during inference of intentions from movement (e.g. Gobbini et al., 2007). The temporal poles on the other hand, which have been associated with memory retrieval, appear to be primarily recruited during the inference of mental states based on a person's previous behavior and attitude (e.g. Ruby and Decety, 2004). Temporal involvement may also partially reflect the importance of listening and language-based processes. The involvement of temporal regions is commonly left lateralized, supporting this assumption, given that language processing primarily relies on the left hemisphere (Adolfi et al., 2017). Somatosensory regions are commonly co-activated with DMN regions during ToM (Atique et al., 2011), which points to the importance of mental simulation to model others' experience, as discussed below in relation to affective processes.

Theory of mind

ToM is the ability to understand and take into account another individual's mental state (Premack and Woodruff, 1978). In humans, ToM and the understanding that a person can hold a false belief develop between the ages of 3–4 years and is fully developed only at the age of 5. O'Connell and Dunbar (2003) studied chimpanzees, a group of autistic children (assumed to lack ToM) and children at ages between 3 and 6 years. “False belief” was tested using nonverbal tests. The chimpanzees performed better than autistic and 3-year-old normal children; they were equal to 4–5-year-old and inferior to 6-year-old children. This would corroborate the idea that chimpanzees exhibit at least some aspects of ToM. At present, the capability of ToM in nonhuman primates remains controversial. Call and Tomasello (2008) report that chimpanzees understand the goals and intentions of others as well as the perception and knowledge of others but found no evidence for understanding false beliefs, while Penn and Povinelli (2007) argue that there is no evidence that nonhuman animals possess anything remotely resembling ToM.

Theory of Mind (ToM)

ToM refers to the capacity of inferring other's mental states (Premack and Woodruff, 1978). It implies coordination of several neural hubs, such as the medial prefrontal cortex (PFC), the bilateral temporo-parietal junction and the medial parietal cortex. Tasks that analyze different ToM components (such as the Reading the Mind in the Eyes Test assessing affective vs the Faux pass tests indexing cognitive ToM) vary in complexity, recruiting different cognitive processes.

ToM is of particular interest in bvFTD, being significantly impaired in this condition (Adolph et al., 2013; Gregory et al., 2002). BvFTD patients suffer systematic deficits in cognitive and affective ToM (Adolph et al., 2013; Gregory et al., 2002). Such deficits possibly reflect impairment in the frontal median network (including the anterior PFC, the ACC, the subcallosal/septal area and the gyrus rectus) in the anterior medial frontal cortex, in Brodmann areas 9 and 32 (Le Bouc et al., 2012), in hypometabolism in the right lateral PFC (Le Bouc et al., 2012), and among perfusion in the left rostral medial PFC (Bertoux et al., 2014). Impairments in ToM tend to be pervasive in other neurodegenerative conditions besides bvFTD, such as in semantic dementia, temporal lobe epilepsy (Hennion et al., 2015), focal brain lesion (Adolfi et al., 2017), Parkinson's Disease (Poletti et al., 2013), AD (Castelli et al., 2011) and stroke patients with temporal and frontal damage (Duval et al., 2012).

Under a dimensional approach, deficits in high-order ToM abilities tend to be more pervasive over those at a basic level. The latter tend to be severely affected in disorders with robust and pronounced social cognition impairments (such as bvFTD) than in those with limited social impairments (such as Parkinson's Disease (PD) and AD). Meanwhile, implicit ToM tasks (Fig. 3) depending on extended neural networks are more sensitive than explicit tasks. These deficits can be explained by several different pathophysiological mechanisms. At the same time, ToM tasks depending on extended neural networks are more sensitive over explicit tasks. These deficits can be explained by several different pathophysiological mechanisms, not only at core but also at unspecific brain disturbances. Thorough research is needed to comprehend the neural underpinnings of different ToM subprocess (emotional, cognitive, high-level ToM, irony, figurative language, perspective taking) and to describe dimensional and transnosological differences across components.

Theoretical considerations

ToM has been conceptualized as a modular and unique aspect of social cognition (Rowe et al., 2001; Havet-Thomassin et al., 2006) that is independent of other cognitive skills. Empirically, however, this is not easy to demonstrate. ToM tasks, by their nature, rely upon multiple cognitive domains such as visual attention and language comprehension for the simplest tasks and executive functioning (flexibility, abstract reasoning), learning, and working memory for more complex tasks. While a few studies have reported no association between performance on ToM tasks and standard neuropsychological tests (Havet-Thomassin et al., 2006; Muller et al., 2010; Spikman et al., 2012), many others have shown a significant correlation between ToM tasks and tests of working memory, processing speed, inhibition, and flexibility (Bibby and McDonald, 2005; Havet-Thomassin et al., 2006; Henry et al., 2006; Milders et al., 2006; Turkstra et al., 2008; Dennis et al., 2009; Channon and Crawford, 2010; Honan et al., 2015). This association could reflect at least partial dependence of ToM ability on these other cognitive functions, or it could reflect the fact that in TBI, the injury commonly affects multiple systems. People with TBI also have difficulty making inferences in general (e.g., inferring the physical cause for an event) (Bibby and McDonald, 2005; Martin and McDonald, 2005; Milders et al., 2006, 2010), which suggests that poor performance on ToM tasks may reflect, at least partially, a general problem with inductive reasoning.

These relationships highlight the multidetermined nature of ToM tasks and therefore raise questions about their specificity. They do not, however, preclude the possibility that specific ToM abilities are disrupted by TBI. Functional neuroimaging studies implicate a specialized neural network that underpins mentalizing, including the bilateral temporoparietal junctions and the anterior dorsal medial prefrontal cortex (Molenberghs et al., 2016). The temporoparietal junction appears to be engaged when inferring intention from purposeful movement (Castelli et al., 2002), while the medial prefrontal cortex is engaged whenever a person thinks about themselves (Northoff et al., 2006) or about others, especially those who are similar to themselves (Mitchell et al., 2005). The pattern of activation of the medial prefrontal cortex during mental tasks is consistent with the notion of simulation, i.e., that people use themselves as a reference point when considering another person's mental state (Gallese and Sinigaglia, 2011). These findings are consistent with the particular vulnerability of medial frontal regions in TBI, which might account for dysfunction in the mentalizing network.

Given that the mentalizing network overlaps considerably with other networks known to underpin executive control, it is possible that ToM deficits in TBI reflect an interaction between executive cognitive impairment and ToM ability. Specifically, if simulation is a mechanism for understanding others, it implies a need to switch between one's own and the other's perspective. Social perspective taking has been found in functional neuroimaging studies to activate inferior dorsolateral and orbitofrontal regions (Ruby and Decety, 2004; D'Argembeau et al., 2007), areas that also mediate cognitive inhibition (Collette et al., 2001). Again, these areas are vulnerable to TBI and inhibitory control is commonly impacted by TBI (Dimoska-Di Marco et al., 2011). Therefore poor inhibition might be expected to affect ToM abilities. Further, this has been supported behaviorally. We found that adults with TBI were relatively good at considering the perspectives of others unless they were first asked to talk about their own point of view. In this situation, they were impaired in their capacity to suppress their egocentric thoughts and instead think about things from another's person's view point (McDonald et al., 2014).

2.7 Theory of mind

Theory of mind (ToM) is defined as the ability to understand and take into account another individual's mental state or of “mind-reading” (Premack and Woodruff, 1978). In children, the ability to implicitly understand the intentions as well as false beliefs of others is already present around the age of 1 year (Baron-Cohen et al., 2013; Meunier, 2017), while “full-blown” or “explicit” ToM and the understanding that a person can hold a false belief develops between the ages of 3 and 4 years and is fully developed only at the age of 5 (Flavell et al., 1978, 1981).

The search for existence of ToM or “mind reading” and understanding false beliefs in monkeys has yielded mixed results. According to Drayton and Santos (2018), rhesus monkeys expect others to update their representations of unseen objects after unsuccessful search. Capuchin, tamarin and macaque monkeys have been reported to understand goal-directed, intentional actions of humans (Schmitt et al., 2012), while in the study of Martin and Santos (2014) monkeys failed. Thus, the presence of ToM in monkeys remains unclear.

For a number of years, the capability of ToM remained controversial even in great apes. Call and Tomasello (2008) reported that chimpanzees understand the goals and intentions of others, as well as the perception and knowledge of others, but found no evidence for understanding false beliefs, while Penn and Povinelli (2007) argued that there is no evidence that non-human animals possess anything remotely resembling ToM including understanding of false belief. However, evidence exist that great apes not only possess a ToM, but also can understand false beliefs, although only at level 1 in the sense of “implicit” perspective-taking (Flavell et al., 1978, 1981), while they fail at explicit level 2. This means that there is no understanding that the same object might appear differently from another perspective (Karg et al., 2016). Children at an age of 2, as well as great apes reveal an implicit understanding of false beliefs, e.g., by spontaneous gaze responses at a violation of their expectations, while explicit level 2 is reached by children only at an age of 4–5 years.

Closely connected to the existence of ToM is the capacity to distinguish between reliable from unreliable information and the formation of trust. This includes habituation to repeated false information (“false alarms”). Chimpanzees understand gaze as a referential information and exhibit habituation to false alarm (Schmid et al., 2017). Children below age 4 discriminate between reliable and unreliable information from a person, and even 14-months-old infants do so at following the experimenter's gaze behind a barrier.