The voluntary control of motor imagery. Imagined movements in individuals with feigned motor impairment and conversion disorder

[Functional neurological disorders and forensic medicine: Clinical and practical characteristics]

Functional neurological disorders (FND) have long been a challenge for medicine, both on clinical and psychodynamic point of view. The medico-legal issue is often relegated to the background in medicine, and FND patients also suffer from such neglected topic. Nonetheless, despite the difficulty to properly diagnose FND and the numerous associated organic and/or psychiatric comorbidities, FND patients report a significant level of deficiency and a high alteration of quality of life when compared to other well-recognized chronic disorders such as Parkinson's disease or epilepsy. Whether it is for the estimation of a personal injury, a prejudice, after-effects following a medical accident or certain legal contexts requiring the elimination of a factitious disorder or a simulation, the uncertainty and imprecision in the medico-legal assessment can imply notable consequences on the patient. In this article, we propose to define the different medico-legal contexts in which FND can occur that of the legal expert, that of the consulting physician, that of the so-called recourse physician and finally that of the attending physician who can provide detailed medical files to the patient in order to help him/her in his/her procedures. We then explain how to use standardized objective evaluation tools validated by the learned societies and how to encourage multidisciplinary cross-evaluation. Finally, we specify how to differentiate FND from historically FND-associated disorders (factitious and simulated disorders), through the clinical criteria, considering the difficulties linked to the uncertainty in the clinical examination of these disorders in a medico-legal context. In addition to the rigorous completion of expertise missions, we aim to reduce two damaging consequences characterizing FND: diagnostic delay and the patients' suffering through stigma.

What Perceivers Know About Their Own Affordance Perception: Post Hoc Evaluations of Perceptual Processes Do Not Relate to Accuracy

In 2 studies, we sought to investigate participants’ awareness of their own cognitive process when perceiving affordances (i.e., action capabilities) and whether their evaluations of this process relate to affordance perception accuracy. In Study 1, we combined data from 8 experiments and created a coding scheme designed to capture the unique features of different affordance perception theories within these post hoc evaluations. We then tested whether the perceptual process that participants reported was related to the accuracy of their judgments. We found that evaluations were distributed fairly evenly across the themes. We also found no relationship between the evaluations and accuracy. In Study 2, we conducted an experiment that specifically addressed our hypotheses about awareness and affordance perception accuracy, while controlling for confounds in Study 1. The greater detail participants gave about their cognitive strategy produced reports that largely did not relate to affordance perception theories. Participants used one of a few strategies aimed at estimating their body dimensions. Our results suggest that awareness of the perceptual process has no influence on the accuracy of affordance perception. It is likely that subjective reports of affordance perceptual process, though clear to the perceiver, are not related to the process itself.

Does Motor Simulation Theory Explain the Cognitive Mechanisms Underlying Motor Imagery? A Critical Review

Motor simulation theory (MST; Jeannerod, 2001) purports to explain how various action-related cognitive states relate to actual motor execution. Specifically, it proposes that motor imagery (MI; imagining an action without executing the movements involved) shares certain mental representations and mechanisms with action execution, and hence, activates similar neural pathways to those elicited during the latter process. Furthermore, MST postulates that MI works by rehearsing neural motor systems off-line via a hypothetical simulation process. In this paper, we review evidence cited in support of MST and evaluate its efficacy in understanding the cognitive mechanisms underlying MI. In doing so, we delineate the precise postulates of simulation theory and clarify relevant terminology. Based on our cognitive-level analysis, we argue firstly that the psychological mechanisms underlying MI are poorly understood and require additional conceptual and empirical analysis. In addition, we identify a number of potentially fruitful lines of inquiry for future investigators of MST and MI.

Neurobiologic theories of functional neurologic disorders

Although neurobiologic theories to explain functional neurologic symptoms have a long history, a relative lack of interest in the 20th century left them far behind neurobiologic understanding of other illness. Here we review the proposals for neurobiologic mechanisms of functional neurologic symptoms that have been made over time and consider how they might inform our diagnostic and treatment methods, and how they integrate with psychologic formulations of functional symptoms. Modern approaches map on to recent developments in theoretic models of brain function, and suggest a key role for processes affecting attention, beliefs/expectations, and a resultant impairment of sense of agency.

Cognitive constraints on motor imagery

Executed bimanual movements are prepared slower when moving to symbolically different than when moving to symbolically same targets and when targets are mapped to target locations in a left/right fashion than when they are mapped in an inner/outer fashion [Weigelt et al. (Psychol Res 71:238-447, 2007)]. We investigated whether these cognitive bimanual coordination constraints are observable in motor imagery. Participants performed fast bimanual reaching movements from start to target buttons. Symbolic target similarity and mapping were manipulated. Participants performed four action conditions: one execution and three imagination conditions. In the latter they indicated starting, ending, or starting and ending of the movement. We measured movement preparation (RT), movement execution (MT) and the combined duration of movement preparation and execution (RTMT). In all action conditions RTs and MTs were longer in movements towards different targets than in movements towards same targets. Further, RTMTs were longer when targets were mapped to target locations in a left/right fashion than when they were mapped in an inner/outer fashion, again in all action conditions. RTMTs in imagination and execution were similar, apart from the imagination condition in which participants indicated the start and the end of the movement. Here MTs, but not RTs, were longer than in the execution condition. In conclusion, cognitive coordination constraints are present in the motor imagery of fast (<1600 ms) bimanual movements. Further, alternations between inhibition and execution may prolong the duration of motor imagery.

Selective Efficacy of Static and Dynamic Imagery in Different States of Physical Fatigue

There is compelling evidence that motor imagery contributes to improved motor performance, and recent work showed that dynamic motor imagery (dMI) might provide additional benefits by comparison with traditional MI practice. However, the efficacy of motor imagery in different states of physical fatigue remains largely unknown, especially as imagery accuracy may be hampered by the physical fatigue states elicited by training. We investigated the effect of static motor imagery (sMI) and dMI on free-throw accuracy in 10 high-level basketball athletes, both in a non-fatigued state (Experiment 1) and immediately after an incremental running test completed until exhaustion (20 m shuttle run-test-Experiment 2). We collected perceived exhaustion and heart rate to quantify the subjective experience of fatigue and energy expenditure. We found that dMI brought better shooting performance than sMI, except when athletes were physically exhausted. These findings shed light on the conditions eliciting optimal use of sMI and dMI. In particular, considering that the current physical state affects body representation, performing dMI under fatigue may result in mismatches between actual and predicted body states.

Disturbed mental imagery of affected body-parts in patients with hysterical conversion paraplegia correlates with pathological limbic activity

Patients with conversion disorder generally suffer from a severe neurological deficit which cannot be attributed to a structural neurological damage. In two patients with acute conversion paraplegia, investigation with functional magnetic resonance imaging (fMRI) showed that the insular cortex, a limbic-related cortex involved in body-representation and subjective emotional experience, was activated not only during attempt to move the paralytic body-parts, but also during mental imagery of their movements. In addition, mental rotation of affected body-parts was found to be disturbed, as compared to unaffected body parts or external objects. fMRI during mental rotation of the paralytic body-part showed an activation of another limbic related region, the anterior cingulate cortex. These data suggest that conversion paraplegia is associated with pathological activity in limbic structures involved in body representation and a deficit in mental processing of the affected body-parts.

Lying in neuropsychology

The issue of lying occurs in neuropsychology especially when examinations are conducted in a forensic context. When a subject intentionally either presents non-existent deficits or exaggerates their severity to obtain financial or material compensation, this behaviour is termed malingering. Malingering is discussed in the general framework of lying in psychology, and the different procedures used by neuropsychologists to evidence a lack of collaboration at examination are briefly presented and discussed. When a lack of collaboration is observed, specific emphasis is placed on the difficulty in unambiguously establishing that this results from the patient's voluntary decision.

Physiological changes in response to apnea impact the timing of motor representations: a preliminary study

Background

Reduced physiological arousal in response to breath-holding affects internal clock processes, leading swimmers to underestimate the time spent under apnea. We investigated whether reduced physiological arousal during static apnea was likely to affect the temporal organization of motor imagery (MI).

Methods

Fourteen inter-regional to national breath-holding athletes mentally and physically performed two 15 m swimming tasks of identical durations. They performed the two sequences in a counterbalanced order, the first while breathing normally using a scuba, the second under apnea. We assessed MI duration immediately after completion of the corresponding task. Athletes performed MI with and without holding breath.

Results

MI durations (26.1 s ± 8.22) were significantly shorter than actual durations (29.7 s ± 7.6) without holding breath. Apnea increased MI durations by 10% (± 5%). Heart rate decrease in response to breath-holding correlated with MI durations increase (p < .01). Under apnea, participants achieved temporal congruence between MI and PP only when performing MI of the apnea swimming task. Self-report data indicated greater ease when MI was performed in a physiological arousal state congruent with that of the corresponding motor task.

Conclusions

Physiological arousal affected the durations of MI through its effects on internal clock processes and by impacting the congruency in physiological body states between overt and covert motor performance. Present findings have potential implications with regards to the possibility of preventing underestimation of durations spent under a state of reduced physiological arousal.

A Bayesian account of 'hysteria'

This article provides a neurobiological account of symptoms that have been called ‘hysterical’, ‘psychogenic’ or ‘medically unexplained’, which we will call functional motor and sensory symptoms. We use a neurobiologically informed model of hierarchical Bayesian inference in the brain to explain functional motor and sensory symptoms in terms of perception and action arising from inference based on prior beliefs and sensory information. This explanation exploits the key balance between prior beliefs and sensory evidence that is mediated by (body focused) attention, symptom expectations, physical and emotional experiences and beliefs about illness. Crucially, this furnishes an explanation at three different levels: (i) underlying neuromodulatory (synaptic) mechanisms; (ii) cognitive and experiential processes (attention and attribution of agency); and (iii) formal computations that underlie perceptual inference (representation of uncertainty or precision). Our explanation involves primary and secondary failures of inference; the primary failure is the (autonomous) emergence of a percept or belief that is held with undue certainty (precision) following top-down attentional modulation of synaptic gain. This belief can constitute a sensory percept (or its absence) or induce movement (or its absence). The secondary failure of inference is when the ensuing percept (and any somatosensory consequences) is falsely inferred to be a symptom to explain why its content was not predicted by the source of attentional modulation. This account accommodates several fundamental observations about functional motor and sensory symptoms, including: (i) their induction and maintenance by attention; (ii) their modification by expectation, prior experience and cultural beliefs and (iii) their involuntary and symptomatic nature.

EEG correlates of Fitts's law during preparation for action

Humans' inability to move fast and accurately at the same time is expressed in Fitts's law. It states that the movement time between targets depends on the index of difficulty, which is a function of the target width and the inter-target distance. The present study investigated the electrophysiological correlates of Fitts's law during action planning using high-density electroencephalography. Movement times were scaled according to Fitts's law, indicating that participants could not overcome the speed-accuracy trade-off during a 1-s preparation period. Importantly, the index of difficulty of the planned movement correlated linearly with the amplitudes of the cognitive N2 and P3b components, which developed during the planning period over parieto-occipital areas. These results suggest that the difficulty of a movement during action planning is represented at a level where perceptual information about the difficulty of the ensuing action is linked to motor programming of the required movement.

Motor imagery may incorporate trial-to-trial error

The authors tested for 1/f noise in motor imagery (MI). Participants pointed and imagined pointing to a single target (Experiment 1), to targets of varied size (Experiment 2), and switched between pointing and grasping (Experiment 3). Experiment 1 showed comparable patterns of serial correlation in actual and imagined movement. Experiment 2 suggested increased correlation for MI and performance with increased task difficulty, perhaps reflecting adaptation to a more complex environment. Experiment 3 suggested a parallel decrease in correlation with task switching, perhaps reflecting discontinuity of mental set. Although present results do not conclusively reveal 1/f fluctuation, the emergent patterns suggest that MI could incorporate trial-to-trial error across a range of constraints.

Mental motor imagery indexes pain: the hand laterality task

Mental motor imagery is subserved by the same cognitive systems that underlie action. In turn, action is informed by the anticipated sensory consequences of movement, including pain. In light of these considerations, one would predict that motor imagery would provide a useful measure pain-related functional interference. We report a study in which 19 patients with chronic musculoskeletal or radiculopathic arm or shoulder pain, 24 subjects with chronic pain not involving the arm/shoulder and 41 normal controls were asked to indicate if a line drawing was a right or left hand. Previous work demonstrated that this task is performed by mental rotation of the subject's hand to match the stimulus. Relative to normal and pain control subjects, arm/shoulder pain subjects were significantly slower for stimuli that required greater amplitude rotations. For the arm/shoulder pain subjects only there was a correlation between degree of slowing and the rating of severity of pain with movement but not the non-specific pain rating. The hand laterality task may supplement the assessment of subjects with chronic arm/shoulder pain.

Mental motor imagery and chronic pain: the foot laterality task

Several lines of evidence suggest that mental motor imagery is subserved by the same cognitive operations and brain structures that underlie action. Additionally, motor imagery is informed by the anticipated sensory consequences of action, including pain. We reasoned that motor imagery could provide a useful measure of chronic leg or foot pain. Forty subjects with leg pain (19 bilateral, 11 right, and 10 left leg pain), 42 subjects with chronic pain not involving the legs, and 38 controls were shown 12 different line drawings of the right or left foot and asked to indicate which foot was depicted. Previous work suggests that subjects perform this task by mentally rotating their foot to match the visually presented stimulus. All groups of subjects were slower and less accurate with stimuli that required a greater degree of mental rotation of their foot. Subjects with leg pain were both slower and less accurate than normal and pain control subjects in responding to drawings of a painful extremity. Furthermore, subjects with leg pain exhibited a significantly greater decrement in performance for stimuli that required larger amplitude mental rotations. These data suggest that motor imagery may provide important insights into the nature of the pain experience.

Fitts's Law violation and motor imagery: are imagined movements truthful or lawful?

Fitts's Law for the timing of targeted movements states that, when target width is held constant, movement time (MT) will increase as the travelled distance increases. Even imagined movements, mentally simulated actions without actual actions, obey Fitts's Law. Recently, a violation of Fitts's Law has been reported; when targets occur in a structured array, MT to the farthest target is shorter than that predicted by Fitts's Law. We conducted two experiments to determine if the violation also occurs for imagined movements. Results showed a close correspondence between real and imaginary MTs across target locations, including the Fitts's violation for the farthest target. We conclude that the violation of Fitts's Law occurs in motor imagery and that the locus of the violation is in the planning stage of action.

Motor inhibition in hysterical conversion paralysis

Brain mechanisms underlying hysterical conversion symptoms are still poorly known. Recent hypotheses suggested that activation of motor pathways might be suppressed by inhibitory signals based on particular emotional situations. To assess motor and inhibitory brain circuits during conversion paralysis, we designed a go-nogo task while a patient underwent functional magnetic resonance imaging (fMRI). Preparatory activation arose in right motor cortex despite left paralysis, indicating preserved motor intentions, but with concomitant increases in vmPFC regions that normally mediate motivational and affective processing. Failure to execute movement on go trials with the affected left hand was associated with activations in precuneus and ventrolateral frontal gyrus. However, right frontal areas normally subserving inhibition were activated by nogo trials for the right (normal) hand, but not during go trials for the left hand (affected by conversion paralysis). By contrast, a group of healthy controls who were asked to feign paralysis showed similar activation on nogo trials and left-go trials with simulated weakness, suggesting that distinct inhibitory mechanisms are implicated in simulation and conversion paralysis. In the patient, right motor cortex also showed enhanced functional connectivity with the posterior cingulate cortex, precuneus, and vmPFC. These results suggest that conversion symptoms do not act through cognitive inhibitory circuits, but involve selective activations in midline brain regions associated with self-related representations and emotion regulation.

The role of motor simulation in action perception: a neuropsychological case study

Research on embodied cognition stresses that bodily and motor processes constrain how we perceive others. Regarding action perception the most prominent hypothesis is that observed actions are matched to the observer's own motor representations. Previous findings demonstrate that the motor laws that constrain one's performance also constrain one's perception of others' actions. The present neuropsychological case study asked whether neurological impairments affect a person's performance and action perception in the same way. The results showed that patient DS, who suffers from a frontal brain lesion, not only ignored target size when performing movements but also when asked to judge whether others can perform the same movements. In other words DS showed the same violation of Fitts's law when performing and observing actions. These results further support the assumption of close perception action links and the assumption that these links recruit predictive mechanisms residing in the motor system.

Action-specific influences on distance perception: a role for motor simulation

Perception is influenced by the perceiver's ability to perform intended actions. For example, when people intend to reach with a tool to targets that are just beyond arm's reach, the targets look closer than when they intend to reach without the tool (J. K. Witt, D. R. Proffitt, & W. Epstein, 2005). This is one of several examples demonstrating that behavioral potential affects perception. However, the action-specific processes that are involved in relating the person's abilities to perception have yet to be explored. Four experiments are presented that implicate motor simulation as a mediator of these effects. When a perceiver intends to perform an action, the perceiver runs a motor simulation of that action. The perceiver's ability to perform the action, as determined by the outcome of the simulation, influences perceived distance.

Motor imagery: a window into the mechanisms and alterations of the motor system

Motor imagery is a widely used paradigm for the study of cognitive aspects of action control, both in the healthy and the pathological brain. In this paper we review how motor imagery research has advanced our knowledge of behavioral and neural aspects of action control, both in healthy subjects and clinical populations. Furthermore, we will illustrate how motor imagery can provide new insights in a poorly understood psychopathological condition: conversion paralysis (CP). We measured behavioral and cerebral responses with functional magnetic resonance imaging (fMRI) in seven CP patients with a lateralized paresis of the arm as they imagined moving the affected or the unaffected hand. Imagined actions were either implicitly induced by the task requirements, or explicitly instructed through verbal instructions. We previously showed that implicitly induced motor imagery of the affected limb leads to larger ventromedial prefrontal responses compared to motor imagery of the unaffected limb. We interpreted this effect in terms of greater self-monitoring of actions during motor imagery of the affected limb. Here, we report new data in support of this interpretation: inducing self-monitoring of actions of both the affected and the unaffected limb (by means of explicitly cued motor imagery) abolishes the activation difference between the affected and the unaffected hand in the ventromedial prefrontal cortex. Our results show that although implicit and explicit motor imagery both entail motor simulations, they differ in terms of the amount of action monitoring they induce. The increased self-monitoring evoked by explicit motor imagery can have profound cerebral consequences in a psychopathological condition.

Motor imagery in physical therapist practice

Motor imagery is the mental representation of movement without any body movement. Abundant evidence on the positive effects of motor imagery practice on motor performance and learning in athletes, people who are healthy, and people with neurological conditions (eg, stroke, spinal cord injury, Parkinson disease) has been published. The purpose of this update is to synthesize the relevant literature about motor imagery in order to facilitate its integration into physical therapist practice. This update also will discuss visual and kinesthetic motor imagery, factors that modify motor imagery practice, the design of motor imagery protocols, and potential applications of motor imagery.

Aiming for the future: prospective action difficulty, prescribed difficulty, and Fitts' law

A main purpose of the current investigation was to determine if Fitts' index of difficulty [log(2)(2A/W)] could be taken as an index of subjective difficulty in prospective action. In two experiments, participants viewed 12 target displays with values of log(2)(2A/W) (prescribed difficulty) ranging between 1.0 and 6.5 bits. Following each 15 s trial, participants provided magnitude estimates reflecting the difficulty that someone else would experience if they actually had to make targeted movements during the trial. The prospective difficulty estimates were always made in the absence of movement. In Experiment 1, target displays were presented to participants on their own video monitor, while in Experiment 2, all participants concurrently viewed scaled-up target displays projected onto a large screen. There were three main findings: First, in both experiments, the prospective-prescribed relation was strong and positive. This finding warrants two conclusions: Fitts' index of difficulty can be taken as an index of subjective difficulty in prospective action, and subjective estimates of performance difficulty result from the monitoring of feedforward control signals generated in the absence of movement-related feedback. Second, even with the large differences in the target display scale of Experiments 1 and 2, difficulty estimates were equivalent at common prescribed difficulty levels. In other words, the results of Experiment 1 were successfully replicated in Experiment 2. This finding demonstrates the generality of the prospective-prescribed relation. Third, nonlinearities in the prospective-prescribed relation resembled those seen in functions describing the increases in movement time that accompany increases in prescribed difficulty (Fitts' law). This observation suggests that the prospective difficulty estimates were based on the value of a temporal parameter in an implicit mental simulation of the action.

Influence of imagined posture and imagery modality on corticospinal excitability

Single pulse transcranial magnetic stimulation (TMS) was used to test the assumption that kinesthetic imagery of action is more 'motor' than visual imagery of action. We assessed corticospinal excitability during motor imagery of a thumb-palm opposition movement by recording potentials evoked by TMS from two hand muscles that would (opponens pollicis, OP, target) or would not (abductor digiti minimi, ADM, control) be activated during actual performance of the very same movement. Participants were asked to imagine the thumb-palm opposition movement while maintaining first person imagery that was either purely visual or predominately kinesthetic. The motor imagery task was performed in two conditions in which the imagined and the actual hand could be either congruent or incongruent. Facilitation of potentials recorded from OP was higher during imagery carried out in mentally congruent than incongruent postures. This effect was largely due to lack of excitability recorded during incongruent kinesthetic imagery, which was indistinguishable from baseline imagery of the static hand. All other conditions differed from static imagery regardless of position. No significant effects were found in a control muscle (ADM) thus indicating that the effect was not related to spatial coding. Subjective reports obtained after the experiment indicate that the results cannot be ascribed to qualitative differences in the imagery experienced. For relatively simple motor tasks requiring no 'expertise' we found no detectable difference in the motor cortex due to imagery modality.

Contribution from neurophysiological and psychological methods to the study of motor imagery

This paper reviews studies on neurophysiological and behavioral methods used to evaluate motor imagery accuracy. These methods can be used when performed in the field and are based on recordings of peripheral indices such as autonomic nervous system or electromyographic activities, mental chronometry and psychological tests. Providing physiological signs that correlate to these types of mental processes may be considered an objective approach for motor imagery analysis. However, although autonomic nervous system activity recording has been shown to match motor imagery in real time, to evaluate its accuracy qualitatively and the individual ability to form mental images, the relationship between physiological responses and mental processes remains an inference. Moreover, electromyographic recordings may be associated with postural control data, but due to inconsistent results, they remain insufficient to solely evaluate motor imagery accuracy. Other techniques traditionally used in psychology and cognitive psychology are questionnaires, "debriefing" with subjects and mental chronometry. Although such methods lead to interesting results, there remains an important part of subjectivity as subjects perform an auto-evaluation of motor imagery accuracy. Similarly, mental chronometry gives information on the ability to preserve temporal organization of movement but does not allow the evaluation of the vividness of mental images. Thus, several methods should be combined to analyze motor imagery accuracy in greater detail. Neurophysiological recordings cannot therefore be considered an alternative but rather a complementary technique to behavioral and psychological methods. The advantages and inconvenient of each technique and the hypotheses that could be tested are discussed.

Perceived reachability: the roles of handedness and hemifield

The impact of reaching experience on the ability to predict one's own reaching range was investigated. Left- and right-handed participants made verbal estimates about the reachability of a target object for both arms in ipsilateral and contralateral frontal space. There was a significant overestimation bias in both groups and for both hands. The overestimation bias increased with the target object's eccentricity in contralateral space. The implications of these findings for models of motor control and motor imagery are discussed.

Abnormalities of motor imagery associated with somatic passivity phenomena in schizophrenia

Some patients with schizophrenia report that their limbs are under the control of an alien force (motor passivity). This is hypothesised to be due to the dysfunction of an internal self-monitoring system that normally permits distinctions between internally generated and external influences on intentional behaviour. Motor imagery is the mental simulation of specific motor actions and it is based upon the internal representation of intended but unexecuted motor actions. Therefore, the generation of motor imagery should be impaired in schizophrenia characterised by passivity phenomena. The generation of motor imagery was compared using the visually guided pointing task (VGPT) and the Florida praxis imagery questionnaire (FPIQ) between patients with schizophrenia characterised by high levels of passivity symptoms (passivity) and patients without passivity symptoms (no-passivity). In both the passivity and no-passivity groups, the speed of real motor sequences on the VGPT was constrained by the distance of the movement and the width of the target in accordance with Fitts' law. For the no-passivity group, the same relationship was found for imagined movements. However, in the passivity group, imagined movements were not constrained by Fitts' law. The effect of a 2-kg load to the limb performing real or imagined movements on the VGPT was identical in both groups. The duration of imagined movements was slowed although the duration of real movements was unaffected. The FPIQ showed that the passivity group had difficulty answering questions that required them to imagine kinaesthetic aspects of performing simple gestures. These results suggest that passivity phenomena in schizophrenia are associated with a specific inability to represent the timing of motor actions internally. This is consistent with the hypothesis that patients with passivity phenomena have difficulty with maintaining an internal representation of intentional behaviour.