**6. Global workspace, social cognition, and their neural substrates in the brain**

For over two decades now, we have been observing the rapid development of research on consciousness [35–37]. This research by addressing questions about subjective nature of experience has resulted in well-empirically established theories that describe formation of conscious knowledge as well as determine directions of contemporary empirical studies on the brain [35, 36]. One of the most well-known accounts on consciousness is a global workspace theory (GWT) postulated by Bernard Baars [33, 38]. The GWT theory has originated in several empirical studies implicating a notion of a neuronal global workspace that has contributed to numerous findings and concepts on possible neural architecture of access consciousness in the brain [39]. The GWT theory has been also found useful in several computational applications, including the field of artificial intelligence or neural network modeling [40].

To sum up, the higher-order structures deal with our rationality which is linked with attribution of available psychological concepts and folk theories to interpret oneself and others' behavior, while lower-level functions of control and monitoring attempt to adjust one's cognitive activity in automatic and unreflective manners [6, 20]. Since the dual-process theory implicates different levels of metacognition, the role of interaction itself is to link such different levels of complexity in a single unit of metacognition [20]. This, in turn, indicates that both levels of metacognition influence one another via such interaction [20], which is critical in dynamic formulating and executing complex cognitive operations in our social responses. Therefore, one can expect that selective impairments in the high-level structures can give rise to impairments in the low-level metacognition and vice versa. For instance, impairments in mindreading processes that develop from childhood onward may cause specific abnormalities in metacognition

Nonetheless, in case of formation of social knowledge predicted by the two-process theory [6], it is important to identify possible mechanisms of accessing the content within such complex capacity. The two-level account proposed by Arango-Muñoz [6] implies that activation of the current content occurs via the interaction which depends on a specific level of processing and a specific psychological context (either related to oneself or other people). This model indicates that higher-order inferential mechanisms should interact with the low-level control and monitoring mechanisms activated in the automatic and unconscious way. Thus, this interactionist account suggests that at the lower level there is no conscious control or monitoring of others' behavior as well as no conscious attribution when handling a specific psychological context of individual either related to the self or others. Clearly, this conclusion is contradicted with common sense that people are often aware of their attributions and can finally formulate accurate and realistic interpretations about others even though they initially produced false attributions. How conscious control or evaluation of social cognition is then possible? It should be emphasized that interactionists (see [6]) claim no explicit architecture of consciousness mechanisms that are responsible for managing the information content within the system. In our opinion, cognitive architecture of such system needs to include a conscious access mechanism that ensures accessibility of information understood as activation through the interaction. Here, we will argue that a combination of the two-level structures of mindreading and metacognition [6] along with a global broadcasting architecture of consciousness [25, 33, 34] is a reasonable theoretical proposal that explains how conscious access contributes in formation of accurate social knowledge.

**6. Global workspace, social cognition, and their neural substrates in** 

For over two decades now, we have been observing the rapid development of research on consciousness [35–37]. This research by addressing questions about subjective nature of experience has resulted in well-empirically established theories that describe formation of

among individuals later in their adulthood diagnosed with schizophrenia [56].

**5. Consciousness as a vehicle of interaction**

**the brain**

110 Prefrontal Cortex

The central claim of GWT is that consciousness has an integrative function that organizes and provides access to a distributed set of knowledge sources that otherwise work as independent structures [25, 34]. According to the conscious access hypothesis, consciousness is considered as an agent that makes the content globally available to unconscious systems [34]. In other words, consciousness enables exchange, coordination, and control of broadcasting the information content among a set of unconscious, specialized, and separate processors [25]. GWT also assumes that the unconscious contents of the mind compete or cooperate with each other in order to gain access to the global workspace. In other words, when the specific information content wins the competition for access over other information, it gets into the neural global workspace that allows its broadcasting to other regions of the brain (specializes processors) in which other processes and resources are activated. In this way, conscious events are results of the interaction between unconscious processors that attempt to spread the information content via the global workspace for other specialized areas of the brain [25, 34, 38].

In the area of brain research, significant progress has been made in understanding the cognitive and neuronal basis of consciousness [41]. Given the cognitive division into conscious and unconscious processing, brain research shows that architecture of consciousness in the brain may be reflected by functionally separate brain regions that are associated with conscious representation and other brain regions responsible for the unconscious processing of lowerorder information to which conscious re-representations are referred [37]. According to the cognitive architecture based on GWT, it is assumed that neural underpinnings of conscious access occur in the prefrontal region (hub) of widely distributed reentrant circuity [41]. Other consciousness studies based on metacognitive approach provide evidence that higher-order representations of consciousness are associated with the activity of prefrontal and parietal cortical structures [42] with a high degree of interconnectivity [43]. It is likely that mechanisms of conscious access localized in the prefrontal and parietal regions receive different kinds of inputs that are required to formulate accurate social interpretation. Following the GWT assumption, unconscious, special-purpose brain processes linked with metacognition and mindreading attempt to get access to a neural global workspace which enables reversible broadcasting to the whole system [44, 45]. Therefore, since mindreading and metacognition constitute unconscious domain-specific processes ("modules"), their neural architecture should be also distinct from conscious structures. For instance, Dimaggio and colleagues [8] show that people who mentalize about themselves (metacognition) and about others (mindreading) activate regions associated with medial prefrontal cortex (mPFC). Interestingly, several regions of the mPFC specialized in social cognition are dissociable when individuals think of others who are perceived as similar or who are dissimilar to the self [8]. Some researchers also suggest that other brain areas such as a bilateral temporal parietal junction (TPJ) may be involved in social cognition as it may be a solid candidate for representing mindreading module [46]. Interestingly, functional neuroimaging studies of clinical population have demonstrated that mindreading deficits are associated with decreasing activation within the medial prefrontal cortex [47]. Some other fMRI study on mindreading deficits in patients with schizophrenia also shows abnormal activation within the left medial prefrontal cortex [48]. Obviously, the actual organization of brain circuity resulting from the proposed framework of linking the dual-process social cognition and the global workspace is considerably more complicated. However, such simplified cognitive architecture of conscious social cognition can allow us to understand how normal and abnormal behavioral and neuronal patterns that accompany conscious processes in social cognition can be developed.

available psychological knowledge. Since conscious access should endorse integrating nature of the metacognition-mindreading relation, its disturbances can lead to abnormal information flow between both subsystems [3]. Here, one can interpret this situation as results of abnormal regulation in accessing the information content that goes in the "from-low-to-highlevel" direction. Therefore, low-level processes of evaluation and monitoring work properly, but the further attribution of psychological contents fails. Thus, our theoretical proposal is that patients with BPD may have disturbances in conscious access that affects activation of interpretative data from metacognition and mindreading to establish proper social behavior. Now, turning to the domain of schizophrenia, we attempt to present how development of clinical symptoms in schizophrenic patients suffering from persecutory delusions can be understood within our proposed framework. With reference to delusional beliefs in schizophrenia, cognitive theories suggest that persecutory delusions often emerge as misinterpretation of social interactions [56]. Therefore, individuals with persecutory delusions are preoccupied with intensions to others [57]. Thus, psychotic patients fail to make accurate judgments in relation to their experiences attributed to others. It has been also suggested that delusional impairments in inferences on the social data may arise from the mindreading deficits. Apparently, deficits in mindreading are demonstrable in schizophrenia as indicated by the meta-analysis by Sprong et al. [21]. Frith [58] hypothesizes that mindreading skills in people with persecutory delusions develop normally; however, those theories of mind capacities are "lost" during psychotic episodes. There is also substantial evidence for mentalizing deficits in patients with first-order episode schizophrenia in the early course of schizophrenia [59]. In fact, mentalizing skills have been shown empirically to be impaired in psychotic patients with persecutory delusions. Patients that follow a paranoid subtype of schizophrenia perform poorly on a wide range of the "theory of mind" tasks including those exercising the attribution of intentions [60]. Moore et al. [61] have explored cognitive etiology of persecutory delusions in patients with late onset of schizophrenia and found that patients performed poorly in a deception task

Consciousness and Social Cognition from an Interactionist Perspective: A New Approach…

http://dx.doi.org/10.5772/intechopen.79584

113

by making more mentalizing errors as compared to healthy participants.

Interestingly, theoretical developments in the conceptualization of relation between metacognition and mindreading skills underlie an interesting casual formation of persecutory delusions. As we mentioned above, the Carruthers' account [4] views metacognition as beliefs of our own attitudes that arise from turning mindreading capacities on ourselves. This implicates that mindreading deficits are prior leading in consequences to dysfunctional metacognition capacities. On the other hand, another explanation is possible as at least deficits in both capacities are paired and may be explained by the interactionist view on social cognition [6]. Indeed, an empirical study by Köther and collaborators [62] on schizophrenic patients could support such interactionist view on persecutory delusions as results of mindreading deficits accompanied with relevant dysfunctional metacognition capacity. In particular, the researchers by employing Reading the Mind in the Eyes test (Eyes Test; [63]) with an additional confidence measure showed in schizophrenic patients not only impaired social cognition in terms of perceiving emotional and social cues but had also commitments to make more high-confidence errors and at the same time made fewer high-confidence correct responses. Obviously, this raises questions about specificity of delusions that may be due to the failures of mentalizing skills and subsequently failures in lower-level metacognition that mirrors
