Elsevier

NeuroImage

Volume 48, Issue 2, 1 November 2009, Pages 486-496
NeuroImage

Neural correlates of processing situational relationships between a part and the whole: An fMRI study

https://doi.org/10.1016/j.neuroimage.2009.06.024Get rights and content

Abstract

Daily situations involve many objects and behaviors. To comprehend the meaning of situations, the relationships between objects, behaviors, and the situational context are important. To reveal the cortical networks involved in processing these relationships we used functional magnetic resonance imaging to compare brain activation during processing of behavior–situation and object–situation relationships. Each session examined two aspects of situational relationship processing: monitoring of the situational relationship and responses to irrelevant relationships. Monitoring was analyzed by comparing cortical activation during a situational relevance judgment task with that during a physical appropriateness judgment task. Responses were analyzed by comparing neural responses to situationally irrelevant and situationally relevant components. The left medial frontal cortex, fusiform gyrus, inferior frontal gyrus, calcarine sulcus, right anterior middle temporal gyrus, orbitoinsular junction, and occipito-temporo-parietal junction were commonly activated while monitoring relationships of both types. The right anterior middle temporal gyrus and orbitoinsular junction were considered to have roles in implicit monitoring because they were more deactivated during physical judgment tasks than during the resting state; this deactivation seemed to reflect unconscious situational monitoring in the resting state. Other regions seemed to be linked to explicit conscious monitoring. Responses to irrelevance were linked to separate and category-specific cortical activation in the left medial frontal cortex and frontal pole for behavioral irrelevance and in the left orbitofrontal cortex for irrelevant objects. We demonstrated that the hierarchical structure of processing situational relationships consisted of implicit monitoring, explicit monitoring, and category-specific responses to irrelevance.

Introduction

A social situation consists of many objects and people behaving variously within it. Both are essential parts of a situation, and the meanings of such situational parts are determined by relationships between the parts and the overall situation. However, no reported study has compared the cortical mechanisms underlying comprehension of the meaning of objects in a situation with those underlying comprehension of the behavior in a situation.

It is obvious that objects are basic parts of a situation and that the relationship between an object and a situation can be understood based on general knowledge. On the other hand, the relationship between a behavior and a situation is considered to be understood by different processes. Behavior–situation relationships are far more complex than object–situation relationships, adding such levels of meaning as morals, manners, and rules. The basis for this complexity may be the role of intention. The cognitive process for comprehension of a behavior–situation relationship is certainly more sophisticated than that for an object–situation relationship.

While there are surely distinctive differences between the cognitive processes of comprehending an object–situation relationship and comprehending a behavior–situation relationship, the existence of general mechanisms for processing the relationship between a part and the whole has been suggested (Jolliffe and Baron-Cohen, 1999, Jolliffe and Baron-Cohen, 2001, Ferstl and von Cramon, 2000, Baumgaertner et al., 2002). This hypothesis was derived from the cognitive characteristics of autistic patients, and has been introduced to explain the pathogenesis of autistic disorder (Jolliffe and Baron-Cohen, 1999, Jolliffe and Baron-Cohen, 2001). Autistic disorder is a neurodevelopmental disorder characterized by social impairment (Kanner, 1943, American Psychiatric Association, 1994). Autistic people tend to focus on details and have difficulty in processing larger entities (Frith, 1989). This tendency has been explained by the hypothesis that they have difficulty processing the relationship between a part and the whole (Frith, 1989, Jolliffe and Baron-Cohen, 1999, Jolliffe and Baron-Cohen, 2001). Many previous studies have shown this cognitive characteristic in autistic patients as evidenced, for example, by difficulty identifying the whole entity from fragments of an object picture, such as puzzle pieces (Jolliffe and Baron-Cohen, 2001), difficulty processing the whole entity of figures (Rinehart et al., 2000), and difficulty finding proper meanings of homographs or ambiguous partial sentences based on the context sentences (Jolliffe and Baron-Cohen, 1999). While impairment in the comprehension of the intentions of another's behavior has been focused on as a core source of difficulty in processing everyday social situations in autistic patients (Baron-Cohen, 1995), the impairment in processing of the part–whole relationship is also considered a reason for the difficulty in situational processing (Stewart, 1995). Impairment in the comprehension of another person's intentions may be explained as impairment in processing the behavior–situation relationship. In contrast, impairment in processing the part–whole relationship cannot be explained as impairment in comprehending another's intentions, because impairment in processing of the part–whole relationship is not limited to the behavioral elements of situations. Impairment in processing the part–whole relationship may explain globally the cognitive impairment seen in the processing of both behavioral and non-behavioral elements in autistic patients.

Although there may be cortical mechanisms common to the cognitive process for behavior–situation relationships and object–situation relationships, each cognitive process has been examined from different viewpoints in previous studies (Berthoz et al., 2004, Bar and Aminoff, 2003, Bar, 2004, Goh et al., 2004, Berthoz et al., 2006, King et al., 2006). Previous studies about processing behavior–situation relationships have focused primarily on perception of the social meanings produced by behaviors in context, such as social transgressions (Berthoz et al., 2004, Berthoz et al., 2006) or social appropriateness (King et al., 2006). These studies have shown the mechanisms for such processes by examining the neural response to situationally inappropriate behaviors. On the other hand, studies about processing object–situation relationships have focused on processing the relationship between an object and a background by examining cortical activation while observing or monitoring the relationship (Bar and Aminoff, 2003, Bar, 2004, Goh et al., 2004). Both phenomena represent necessary aspects of situational processing. Situational monitoring is likely to contribute to the comprehension of situations by functioning in any state, in that we are awake both explicitly and implicitly. Neural responses to situational inappropriateness may contribute to the identification of the situational meanings of situationally irrelevant parts that might carry significant social meaning and might require immediate behavioral responses, such as imminent danger or an unexpected reward. Previous studies have dealt with situational monitoring as an intentional cognitive process (Bar and Aminoff, 2003, Bar, 2004, Goh et al., 2004). Thus, cortical activation while monitoring a situational relationship chiefly depends on intentional cognitive processing of the relationship between parts and a situation, as shown in previous studies. This form of processing relies more on top-down processing. The neural response to a situationally inappropriate element is driven by the relationship between a salient stimulus and a situation. Thus, this form of processing relies more on bottom-up and transient processing than does the monitoring of situational relationships. Addressing these different foci, previous studies about behavior–situation relationships and object–situation relationships have identified separate neural networks for the two. For example, the medial prefrontal regions and the amygdala have been detected in the processing of behavior–situation relationships, and the parahippocampal area, the fusiform gyrus, and the retrosplenial cortex have been implicated in the processing of object–situation relationships (Berthoz et al., 2004, Bar and Aminoff, 2003, Bar, 2004, Goh et al., 2004, Berthoz et al., 2006, King et al., 2006).

We sought to find a general neural network for processing situational relationships between parts and the whole, and further, to determine the entire structure for processing situational relationships. To find general neural mechanisms for processing part–whole relationships in situations, it is necessary to investigate both types of elements, behaviors and objects, using the same conceptual bases; the monitoring of situational relationships and the response to situationally inappropriate parts. Cortical activation in the monitoring of relationships and in the response to inappropriateness were both examined to analyze behavior–situation and object–situation relationships, and the networks for processing behavior–situation and object–situation relationships were compared. Comparing the cortical activation involved in monitoring a situational relationship with the neural response to a situationally inappropriate element may reveal how each contributes to the general process of responding to situational relationships between parts and the whole, and how each constructs the whole structure for situational processing.

In this study, pictures of everyday behaviors and objects were prepared as parts stimuli, and pictures of everyday situations were prepared as whole stimuli. Behavior–situation and object–situation relationships were examined in separate sessions. In two sessions, each healthy adult subject was asked to perform two types of tasks while being scanned by functional magnetic resonance imaging (fMRI). In the first, the subject was instructed to detect the situationally irrelevant part in a contextual situation. In the second task, the control task, the subject was instructed to detect the vertically inverted part in a contextual situation. The control task was designed to evoke the cognitive process of simple error detection, visual input, and motor response. Our special interest is the specific neural activation involved in the process of monitoring situational relationships or in the response to irrelevance. Thus, neural activation for low-level cognitive processes, such as visual input and motor responses, should be excluded from the monitoring process, and non-specific simple error detection from the response to irrelevance. To investigate the cognitive process for monitoring situational relationships, neural activation during the situational relevance judgment task was compared with that during the control task. To examine the neural response to inappropriateness of a situational relationship, we compared the neural response to detecting irrelevant part stimuli against whole stimuli with the response to detecting relevant part stimuli against whole stimuli.

We also tried to differentiate between the implicit and the explicit aspects of the monitoring process; previous studies have not separated the monitoring process into explicit conscious and implicit unconscious processes (Bar and Aminoff, 2003, Bar, 2004, Goh et al., 2004). We hypothesized that the cognitive process for monitoring situational relationships is implicitly involved in any waking state, including the resting state during fMRI scanning, and is deactivated during attention-demanding tasks unrelated to situational monitoring. Previous studies have indicated that attention-demanding tasks decreased cortical activation for processing related to other sensory modalities (Just et al., 2008, Mazolic et al., 2008). Those results suggested that attention demand suppressed other processes that were unrelated to the operating process and that such suppression was expressed as cortical deactivation. Other, previous studies have shown that attention-demanding tasks decreased cortical activation during resting states (Shulman et al., 1997, Gusnard and Raichle, 2001, Mazoyer et al., 2000, Raichle et al., 2001, McKiernan et al., 2003, Iacoboni et al., 2004). These studies have been understood as showing that attention demand during resting states suppressed some types of sustained processes that were unrelated to the process underlying the explicit task. Thus, we examined regions that were deactivated in the control task to clarify both implicit unconscious monitoring and explicit conscious monitoring processes. Cortical activity for the implicit monitoring of situational relationships should be higher during the resting state than during the control task because the control task demands attention but is unrelated to situational monitoring. On the other hand, explicit conscious monitoring processes should not be suppressed during the control task.

Section snippets

Subjects

Forty-five healthy right-handed Japanese volunteers (33 males, 12 females, aged 18–23 years, mean 20.8) participated. All subjects had normal vision and none had a history of neurological or psychiatric illness. Handedness was evaluated using the Edinburgh Handedness Inventory (Oldfield, 1997). Written informed consent was obtained from all subjects according to the guidelines of the ethical committee of Tohoku University and the Declaration of Helsinki (1991).

Stimuli and tasks

Pictures depicting an everyday

Behavioral data

The mean reaction time and the mean percentage of correct responses for each trial type are shown in Table 3. Two-way ANOVAs (stimulus type × subject in each session) revealed significant main effects of trial type on both the reaction time and the percentage of correct responses (reaction time in the Behavior session: F5,170 = 85.6, p < 0.001, percentage of correct responses in the Behavior session: F5,170 = 36.70 p < 0.001, reaction time in the Object session: F5,170 = 35.70, p < 0.001, percentage of

Discussion

We demonstrated distinct neural networks for monitoring situational relationships and for responses to irrelevant situational relationships. Neural activation underlying processes for monitoring situational relationships was constructed as a common neural network between those for behavior–situation and object–situation relationships. However, neural networks for responses to irrelevant relationships were separate networks, according to the category of the part (object vs. behavior). Next, we

Conclusions

We found a general neural network for processing situational relationships between parts and the whole, and demonstrated the entire structure for processing situational relationships. Many cortical regions were identified in common between behavior–situation monitoring and object–situation monitoring, and these regions were considered to be a general neural network for processing situational relationships between parts and the whole. Of these regions, the right anterior middle temporal gyrus

Acknowledgments

We thank Atsushi Sekiguchi (IDAC, Sendai, Japan) and Ai Fukushima (IDAC, Sendai, Japan) for support in stimulus preparation. This study was supported by RISTEX/JST, CREST/JST, and the 21st Century Center of Excellence (COE) Program (Ministry of Education, Culture, Sports, Science and Technology) entitled “A Strategic Research and Education Center for an Integrated Approach to Language, Brain and Cognition” (Tohoku University). Motoaki Sugiura is supported by a Grant-in-Aid for Young Scientists

References (51)

  • LévesqueJ. et al.

    Neural correlates of sad feelings in healthy girls

    Neuroscience

    (2003)
  • NoppeneyU. et al.

    The neural areas that control the retrieval and selection of semantics

    Neuropsychologia

    (2004)
  • OhiraH. et al.

    Association of neural and physiological responses during voluntary emotion suppression

    NeuroImage

    (2006)
  • SaxeR. et al.

    Making sense of another mind: the role of the right temporo-parietal junction

    Neuropsychologia

    (2005)
  • WagnerA.D. et al.

    Recovering meaning: left prefrontal cortex guides controlled semantic retrieval

    Neuron

    (2001)
  • WakusawaK. et al.

    Comprehension of implicit meanings in social situations involving irony: a functional MRI study

    NeuroImage

    (2007)
  • Diagnostic and Statistical Manual of Mental Disorders, 4th ed

    (1994)
  • AmodioD.M. et al.

    Meeting of minds: the medial frontal cortex and social cognition

    Nat. Rev. Neurosci.

    (2006)
  • BarM.

    Visual objects in context

    Nat. Rev. Neurosci.

    (2004)
  • Baron-CohenS.

    Mindblindness: An Essay on Autism and Theory of Mind

    (1995)
  • BartelsA. et al.

    The neural basis of romantic love

    Neuroreport

    (2000)
  • BerthozS. et al.

    An fMRI study of intentional and unintentional (embarrassing) violations of social norms

    Brain

    (2004)
  • BoxterA.L. et al.

    Clinical features of fronttemporal dementia

    Alzheimer Dis. Assoc. Disord.

    (2005)
  • DeLongG.R. et al.

    Acquired reversible autistic syndrome in acute encephalopathic illness in children

    Arch. Neurol.

    (1981)
  • EslingerP.J. et al.

    Oops! Resolving social dilemmas in frontotemporal dementia

    J. Neurol. Neurosurg. Psychiatry

    (2007)
  • Cited by (7)

    • Adaptive ability to cope with atypical or novel situations involving tool use: An fMRI approach

      2015, Neuroscience Research
      Citation Excerpt :

      It is possible that the greater emphasis placed on situational context in this study caused these regions to be activated to a greater degree than in previous studies. Previous fMRI studies have linked the left lateral orbitofrontal cortex, the bilateral insula, and the dorsal medial frontal cortex to the processing of the adaptive relationship between behaviors and social situations (Marshal et al., 2005; Sassa et al., 2007; Wakusawa et al., 2009). Activation in the left inferior temporal gyrus and the dorsal anterior cingulate cortex may reflect a high load of referring semantics employed to identify the adaptive meanings of novel or atypical tools in situational contexts; these two regions were reported to be linked to semantic processing (Creem-Regeher and Lee, 2005; Harrington et al., 2009) and be involved in action production during tool use by fMRI studies (Johnson-Frey et al., 2004; Ruminati et al., 2004).

    • Beyond human intentions and emotions

      2013, Frontiers in Human Neuroscience
    View all citing articles on Scopus
    View full text