The Neural Dynamics Of Conflict Adaptation

To optimize a goal-directed performance, humans usually need to accomplish current task according to previous experience. When confronting with difficulties or interference, the brains can effectively utilize cognitive resources to resolve the difficulties or suppress interference on the basis of the previous experience. Accordingly, human current performances are optimized (Miller&Cohen,2001). This phenomenon is named as conflict adaptation in the study of cognitive control (Gratton, Coles,&Donchin,1992), which was first observed by Gratton et al. in the flanker task (Eriksen&Eriksen,1974). Henceforth, some investgaters observed significant conflict adaptation in other interference tasks (Egner,2007). Currently, two dominant theories can be used to explain conflict adaptation:the conflict monitoring theory (Botvinick, Braver, Barch, Carter,&Cohen,2001) and the feature integration theory (Mayr, Awh,&Laurey,2003). The conflict monitoring theory proposes that conflict adaptation consists of two basic courses: conflict detection and top-down control modulation. However, the antergic feature integration theory proposes that conflict dapatation results from bottom-up feature repetition priming. Recently, although both of the theories can be supported by respective experimental data, the neural mechanisms of conflict adaptation remain controversial.Although conflict adaptation is affected by feature repetition priming, which can be eliminated by experimental manipulation, the responses between sequential trials are always "transformational", which may confound response switching effect (Franke, Reuter, Breddin,&Kathmann,2009; Kenner et al.,2010; Reuter, Philipp, Koch,&Kathmann,2006). Compared with congruent condition, the response switching produce a more effect on the incongruent trials, eliminating the feature repetitional trials will be disadvantageous of the conflict resolution in the current trials, which therefore covers conflict adaptation. Thereby, to observe "pure" conflict adaptation, investigators should control the repetition priming effect, negative priming effect and the response switching effect. On the basis of foregoing analyses, we (Tang, Liu,&Chen,2012) designed a conflict observation paradigm based on the color-word Stroop task. Two types of tasks are included in the paradigm:observation of conflict (look at the color of the Stroop words, but no responses are required) and manual response task (respond to the color of the Stroop words). The results revealed that a reliable conflict adaptation was obtained in the manual response task, indicating that the observation of conflict can induce the top-down conflict adaptation.However, that is a behavioral study, the neuromechanisms of conflict adaptation can not be investigated. Recently, we adopted the event-related potential (ERP) technology, with high temporal resolution, and designed the conflict observation paradigm based on four-reaction-set color-word Stroop task and letter flanker task, where the stimulus-response repetitions and the proportion of congruent/incongruent condition were controlled (Tang, Hu, Chen,&Antao,2013; Tang, Hu, Li, Zhang,&Chen,2013; Tang&Chen,2012). The results showed reliable conflict adaptation both in the behavioral and ERP data. However, ERP is normally obtained from the across-trial averaging of several electroencephalographic (EEG) traces, and the obtained ERP waveforms are time-and phase-locked to the onset of a stimulus (e.g., a visual stimulus) or a response (e.g., pressing a response button)(Cohen,2011). In addition to the phase-locked ERP, the presentation of a stimulus would be able to induce non-phase-locked modulations of ongoing oscillatory EEG activity, which can not be detected using the time-domain across-trial averaging (Makeig, Debener, Onton,&Delorme,2004).Recently, using the time-frequency analysis, the obtained data can show multiple-domain EEG information in the time-frequency domain, including time, frequency (the speed of the oscillation), power (the amount of the energy in a frequency band at a point in time), and phase angle (the position of the oscillation along the sine wave, driven by the state of excitation of the population of neurons)(Cohen,2011; Makeig, Debener, Onton,&Delorme,2004). The information related to cognitive processes usually is reflected in specific frequency range, consists of either a transient increase or decrease (event-related synchronization or desynchronization, ERS or ERD) of EEG power. Depending on their frequencies, they may represent neuronal mechanisms involved in cortical activation, inhibition and binding (Pfurtscheller&Lopes da Silva,1999).As been revealed by the conflict monitoring theory, conflict adaptation involves conflict-induced sequential regulation of cognition, which is reflected in the top-down information transfer between the anterior cingulate cortex (ACC) and the dorsal lateral prefronal cortex (DLPFC). That is, the ACC detects the occurrence of conflict and signals the DLPFC, whch modulates the attentional resources and implements the control of conflict. However, existing researches have not directly investigated the direction and strength of such information transfer using the EEG technology. Recently, effective connectivity analysis has been wildly applied to the cognitive researches (e.g., Nigbur, Cohen, Ridderinkhof,&Sturmer,2011; Hanslmayr, Gross, Klimesch,&Shapiro,2011; van Wijk, Daffertshofer, Roach,&Praamstra,2009). It bases on the concept of Granger causality (Granger,1969) and has demonstrated a powerful capacity of evaluating the direction and strength of causality between neuronal activations (Hu, Zhang,&Hu,2012). Therefore, some direct evidence for the process of evaluating conflict and implementing control can been obtained by adopting the effective connectivity analysis, the results of which can powerfully expand the conflict monitoring theory. In the present study, the behavioral and EEG data were recorded from fifteen healthy human subjects when they performed a letter flanker task within the conflict observation paradigm. The data were analyzed both in behavior (response time, RT; error rates) and in electrophysiology (time-domain and time-frequency-domain). For RT, the results showed a significant interaction between the congruency in the observation condition and the congruency in the response condition. In the time domain, the ERP results showed that in the observation task, the mean amplitude of N450(400-450ms; centro-frontal region) was more negative in incongruent condition compared to congruent condition. In the response task, a significant interaction between the congruency of observation condition and the congruency of reponse condition was found for the sustained potential (SP,750-800ms, left-frontal region) amplitude; the amplitude difference was significantly smaller for SP(il-ic) than for SP(cl-cC); and a significant negative correlation between the SP(cI-cC)-(iI-iC) and the RT(cI-cc)-(iI-iC) was observed. These results confirm that in the observation task, the modulations of N450amplitude reflect the process of conflict detection; in the response task, the modulations of SP amplitude reflect conflict adaptation that is induced by the observation of conflict.In the time-frequency domain, the results reveal some more comprehensive and strong evidence for the overall process of conflict adaptation. Time-frequency analysis showed the following findings. In the observation task, both alpha-band (9-13Hz,480-980ms) and beta-band (13-25Hz,480-780ms) ERD in the left-and centro-frontal regions were observed in incongruent condition, but alpha-band and beta-band ERS in these regions were observed in congruent condition. These results indicate that the conflict in incongruent condition induces the modulations of brain in the special frequency regions (alpha-and beta-band) and results in that more cognitive resources are put into controlling of the conflict, effectivelly. In the response task, the magnitude of theta-band (6-8Hz,50-1,000ms) ERS in the left-, right-, and centro-frontal regions showed significant interaction between the congruency in the observation task and the congruency in the response task. Specially, theta-band ERS in the left-and centro-frontal regions were significantly stronger in cl condition than in il condition. The results indicate that the conflict in the response task induces the magnitude modulations of brain oscillatory in theta band, which reflects conflict adaptation. Effective connectivity analysis showed some novel findings. In the observation task, a stronger information flow from right-frontal region to centro-frontal region in the beta-band (26-27Hz,200-1,000ms) was observed for congruent condition than for incongruent condition. That indicates that the information flow at beta-band activities may be related to the conflict information (conflict-related information or congruent information). In the response task, a stronger information flow from right-frontal region to centro-frontal region in the beta-band (22-25Hz,200-1,000ms) was observed for iI condition than for cl condition. That indicates that the congruency context in the observation task affects the strength of information transfer in the response task. The sequential conflict context (iI condition) induces the sequential modulations of conflict in humun brain, and therefore the degree of control on distraction reaction is higher.In conclusion, the present study creatively integrated the conflict observation paradigm with letter flanker task, and investigated the neural mechanisms of conflict adaptation in depth. First, the present study is conducive for investigators to effectively separate the processes of conflict detection from conflict resolution, further understand the time course of conflict adaptation; meanwhile, it will be conducive for us to expand the conflict monitoring theory. Second, the present study not only makes an important contribution to improving the tranditional experiment paradigm, but also adopts a multi-angle EEG data analysis methods and reveals more overall information related to coflict adaptation. In the future studies, investigators can integrate EEG with functional magnetic resonance imaging (fMRI) to study the space-time dynamics of conflict adaptation or use other conflict tasks, such as Simon task (Simon,1969) to expand the present results.