Dynamic goal states: Adjusting cognitive control without conflict monitoring
Highlights
► We present a computational model of cognitive control processes. ► The model uses network dynamics instead of a specialized conflict monitoring module. ► The model mirrors behavioral conflict adaptation and EEG data. ► It mirrors neural correlations of previous imaging studies between ACC and PFC. ► A special conflict monitoring module is not necessary for conflict adaptation.
Introduction
While on your way to the kitchen to get a knife, coming across the watering can you start to water your plants, wondering later what you originally intended to do in the kitchen. This everyday example illustrates the importance of cognitive control processes when we pursue goals in environments that contain abundant distracting affordances. Focusing on goals is necessary to perform non-routine tasks successfully, especially in the face of competition from prepotent habitual or stimulus-triggered responses. In experimental psychology, the context-sensitive recruitment of cognitive control processes has often been studied by using conflict tasks.
In such conflict tasks, the basic recruitment of cognitive control processes is measured by the size of the ‘congruency effect’, that is, the difference in response time (RT) between incongruent (response conflict) and congruent (no response conflict) trials. This congruency effect is supposed to reflect the additional time necessary to resolve increased response conflict in incongruent as compared to congruent trials. For instance, in the Stroop task (Stroop, 1935) participants have to name the ink color of a color word (e.g. ‘red’) while ignoring the word meaning (e.g. ‘blue’). When the color and the word meaning are incongruent, RT is increased compared to trials on which color and word meaning are congruent. Similarly, in the flanker task (Eriksen and Eriksen, 1974), participants have to respond to a centrally presented target (e.g. letters), but have to ignore surrounding flanker stimuli, which typically produce increased RT when the response indicated by the target and the response indicated by the flanker stimuli are incongruent.
Importantly, for the mentioned conflict tasks it has been found that the congruency effect (i.e., reflecting additional time needed to resolve conflict) can be modulated by additional context variables implicating processes of control adjustments working on different time scales. On a short time scale, conflict in the previous trial modulates the recruitment of control, leading to a reduced congruency effect following incongruent trials (Gratton et al., 1992). Hence, this ‘conflict adaptation effect’ could reflect a process that adjusts control from trial to trial. On a longer time scale, the overall frequency of incongruent trials modulates the basic recruitment of control, leading to an increased congruency effect for a lower overall frequency of incongruent trials (Gratton et al., 1992, Tzelgov et al., 1992). Hence, this ‘congruency proportion effect’, could reflect a process that adjusts control across a series of trials.
Section snippets
Conflict monitoring theory
A major step towards a mechanistic explanation of the context-sensitive regulation of cognitive control was taken when Botvinick et al. (2001) presented the conflict monitoring theory, which is still one of the most influential attempts to account for flexible adjustments of cognitive control. This theory offered a powerful and parsimonious account of the context-sensitive regulation of cognitive control with its central assumption that a specific neural module, localized within the anterior
Continuous conflict adaptation without conflict monitoring
In this article, we propose an alternative neural network model that integrates the two views with the original idea of conflict monitoring theory, namely that conflict is responsible for increases of cognitive control. However, instead of an explicit conflict detection and signaling module causing these increases, we propose that context-sensitive adjustments of cognitive control occur within trials (see also Mayr and Awh, 2009), as an emergent by-product of the interaction dynamics within the
Layers
An outline of our model can be seen in Fig. 1A (for details, please see Appendix I). Similar to the structure of previous models solving conflict tasks (Botvinick et al., 2001, Cohen et al., 1992, Gilbert and Shallice, 2002), the model contains two input layers, a response layer and a task/goal layer.
The units in the input layers represent the relevant and irrelevant information in a conflict task, e.g. color and word meaning in a Stroop task. The output layer contains units indicating the
Discussion
We presented a dynamic connectionist model as an account of how cognitive control could be adjusted continuously within trials and without building on an explicit conflict monitoring module. As a proof of concept, the model successfully reproduces standard effects indicating the adaptation of control across trials, i.e. conflict adaptation and the proportion congruent effect. Critically, the model was able to account for these effects without the assumption of an explicit conflict monitoring
Acknowledgment
This research was partly supported by the German Research Council (DFG) (grant SFB 940/1 2012).
References (72)
- et al.
Conflict monitoring and anterior cingulate cortex: an update
Trends Cogn. Sci.
(2004) - et al.
Cortical coordination dynamics and cognition
Trends Cogn. Sci.
(2001) Conflict effects without conflict in anterior cingulate cortex: multiple response effects and context specific representations
NeuroImage
(2009)- et al.
A computational model of fractionated conflict-control mechanisms in task-switching
Cogn. Psychol.
(2007) - et al.
Persistent activity in the prefrontal cortex during working memory
Trends in Cognitive Sciences
(2003) A computational study of conflict-monitoring at two levels of processing: reaction time distributional analyses and hemodynamic responses
Brain Res.
(2008)- et al.
A model of dual control mechanisms through anterior cingulate and prefrontal cortex interactions
Neurocomputing
(2006) - et al.
The neural correlates and functional integration of cognitive control in a Stroop task
NeuroImage
(2005) - et al.
Attentional bias of competitive interactions in neuronal networks of early visual processing in the human brain
NeuroImage
(2008) - et al.
Sustained vs. transient cognitive control: evidence of a behavioral dissociation
Cognition
(2010)
The prefrontal cortex — an update: time is of the essence
Neuron
Task switching: a PDP model
Cogn. Psychol.
The dorsal medial frontal cortex is sensitive to time on task, not response conflict or error likelihood
NeuroImage
Moving to higher ground: the dynamic field theory and the dynamics of visual cognition
New Ideas in Psychology
Practice-related effects demonstrate complementary roles of anterior cingulate and prefrontal cortices in attentional control
NeuroImage
Action sets and decisions in the medial frontal cortex
Trends Cogn. Sci.
How decisions evolve: the temporal dynamics of action selection
Cognition
The anterior cingulate as a conflict monitor: fMRI and ERP studies
Physiol. Behav.
The role of prefrontal regions in the Stroop task
Neuropsychologia
Cognitive conflict without explicit conflict monitoring in a dynamical agent
Neural Netw.
Errors of interpretation and modeling: a reply to Grinband et al
NeuroImage
Dynamics of pattern formation in lateral-inhibition type neural fields
Biological Cybernetics
Item-specific adaptation and the conflict-monitoring hypothesis: a computational model
Psychol. Rev.
Conflict monitoring and decision making: reconciling two perspectives on anterior cingulate function
Cogn. Affect. Behav. Neurosci.
Conflict monitoring and cognitive control
Psychol. Rev.
Learned predictions of error likelihood in the anterior cingulate cortex
Science
Executive control in the Simon effect: an electromyographic and distributional analysis
Psychol. Res.
Error negativity does not reflect conflict: a reappraisal of conflict monitoring and anterior cingulate cortex activity
J. Cogn. Neurosci.
A parallel distributed processing approach to automaticity
Am. J. Psychol.
If it's hard to read… try harder! Processing fluency as signal for effort adjustments
Psychol. Res.
Going, going, gone: characterizing the time-course of congruency sequence effects
Front. Cogn.
Behavioral conflict, anterior cingulate cortex, and experiment duration: implications of diverging data
Hum. Brain Mapp.
Effects of noise letters upon the identification of a target letter in a nonsearch task
Percept. Psychophys.
Dynamic field theory of movement preparation
Psychol. Rev.
Is anterior cingulate cortex necessary for cognitive control?
Brain
Context-sensitive adjustments of cognitive control: conflict-adaptation effects are modulated by processing demands of the ongoing task
J. Exp. Psychol. Learn. Mem. Cogn.
Cited by (31)
One cannot simply 'be flexible’: regulating control parameters requires learning
2024, Current Opinion in Behavioral SciencesCognitive effort and active inference
2023, NeuropsychologiaChoice perseveration in value-based decision making: The impact of inter-trial interval and mood
2019, Acta PsychologicaCitation Excerpt :Taking this as a given, we will focus is on the process dynamics of how the decisions unfold over time (Spivey & Dale, 2004, 2006; van Rooij, Favela, Malone, & Richardson, 2013). In the model, options are represented as self-sustainable active patterns that constitute attractors in the neural system's state space (Reilly, 2006; Rolls, 2010; Scherbaum et al., 2016; Scherbaum, Dshemuchadse, Ruge, & Goschke, 2012). The dynamics of such models are often illustrated by conceiving of the system's state space in analogy to a potential landscape, where the stability of a state depends on the depth of the attractor (Rolls, 2010; Scherbaum, Dshemuchadse, & Kalis, 2008).
Development of between-trial response strategy adjustments in a continuous action control task: A cross-sectional study
2017, Journal of Experimental Child PsychologyCitation Excerpt :It is even possible that control settings are carried over. For example, Scherbaum and colleagues argued that resolution of conflict within a trial requires changes in attentional control settings and that between-trial effects are merely the result of a carryover of these within-trial adjustments (Scherbaum, Dshemuchadse, Ruge, & Goschke, 2012; Scherbaum, Fischer, Dshemuchadse, & Goschke, 2011; see also, e.g., Blais & Verguts, 2012). In other literatures, it has also been argued that lingering effects of past experience can have a large impact on performance (e.g., Awh, Belopolsky, & Theeuwes, 2012).
Age-related changes in strategic variations during arithmetic problem solving: The role of executive control
2016, Progress in Brain ResearchCitation Excerpt :Congruency sequence effects refer to reduced congruency effects on a given item after an incongruent item relative to following a congruent item. These effects were interpreted as resulting from trial-to-trial modulations of top-down control processes (Botvinick et al., 2001; De Pisapia and Braver, 2006; see Duthoo et al., 2014; Scherbaum et al., 2012 for alternative views). Finding similar patterns of sequential modulations in both arithmetic and conflict tasks suggests similar sequential modulations of executive control processes involved.