States of focused attention and sequential action: A comparison of single session meditation and computerised attention task influences on top-down control during sequence learning☆
Introduction
Motor sequences are an integral part of everyday life. Activities of daily living such as driving to work, typing up documents, or preparing a meal, allow us to navigate and interact with the environment successfully. Although these motor sequences are performed with automaticity, cognitive research has yet to provide clear theoretical perspectives on information processing and sequence learning strategies. Recently, cognitive processes utilised in goal-directed behaviours are thought to play a crucial role in sequence learning which leads to the idea that movement sequences can be executed using different learning strategies (Verwey, Shea, & Wright, 2015). For example, cognitive control is a key component in sequence learning and governs how attention is utilised in goal-directed behaviours. This raises the possibility that factors which influence cognitive control may also impact motor sequence learning. Of recent interest is the influence of meditation on cognitive control and especially attention, for subsequent cognitive processes utilised in various goal-directed behaviours. While different forms of meditation exist (Nash & Newberg, 2013), focused attention meditation (FAM) in particular appears to influence attentional control processes (Lutz, Slagter, Dunne, & Davidson, 2008). FAM is characterised by maintaining sustained attention on a specific instructed object (e.g. breath or body awareness) (Lutz et al., 2008; Slagter, Davidson, & Lutz, 2011). It has been shown to constrain attention in a narrow manner which, in turn, bias cognitive control to function in a convergent style for subsequent cognitive tasks (Colzato, Sellaro, Samara, Baas, & Hommel, 2015; Colzato, Sellaro, Samara, & Hommel, 2015). The cognitive control effects of FAM have recently been shown to extend to complex sequence learning (Chan, Immink, & Lushington, 2017; Immink, Colzato, Stolte, & Hommel, 2017). Theoretical models of FAM (Malinowski, 2013; Tang, Holzel, & Posner, 2015) have proposed that as an outcome of practice, core regulatory processes such as effort (Immink et al., 2017; Lumma, Kok, & Singer, 2015), arousal (Amihai & Kozhevnikov, 2015), and pleasure may have an influence on cognitive control but remain poorly understood. We provide a brief review of evidence for FAM on cognitive control and models of sequence learning control strategies.
Motor sequence learning research has provided several cognitive control paradigms that explain the acquisition and representation of sequential action (Abrahamse & Noordzij, 2011; Verwey et al., 2015; Verwey & Wright, 2014). Common amongst these are the interactions between attention, memory, executive functions, and the development of models for task representation, information processing and error resolution for learning improvements over time (Abrahamse, Jimenez, Verwey, & Clegg, 2010; Abrahamse & Noordzij, 2011; Daltrozzo & Conway, 2014; Keele, Ivry, Mayr, Hazeltine, & Heuer, 2003; Verwey et al., 2015). These models converge onto the evidence for two different learning strategies that stem from cognitive control (Verwey & Wright, 2014), namely responding to stimuli in an external stimulus-based control or using sequence-specific representations via an internal plan-based control to engage in sequence learning (Tubau, Hommel, & Lopez-Moliner, 2007; Verwey et al., 2015; Verwey & Clegg, 2005). Importantly, these two strategies demonstrate that movement sequences can be executed with different processing strategies, which signifies that sequence learning is a cognitive task that relies on both central and perceptual processes (Verwey et al., 2015). This highlights the crucial role of cognitive control during learning (Tubau et al., 2007), where factors prior to and during learning can often bias cognitive control and therefore learning strategy.
Although both stimulus- and plan-based control strategies predict sequence learning improvements with practice, the processes underlying performance improvement differ. This can be demonstrated using the motor learning paradigm known as the Serial Reaction Timed Task (SRTT: Nissen & Bullemer, 1987). In the SRTT, a stimulus appears at one of several locations on a screen and participants must respond by pressing the corresponding key according to the stimulus location on the screen. Unbeknownst to them, the order of the stimuli follows a structured sequence that repeats over cycles. Typically after several blocks of training, a transfer block with either a new or random sequence is presented and participants' exhibit increased reaction times due to learning of the original sequence that has occurred over the training blocks. The SRTT can be used to discern which strategy was utilised during learning. For example if an individual is using stimulus-based control, response latency decreases can be explained by the reinforcement of stimulus-response associations due to enhanced top-down cognitive control resulting in the prioritisation of attention at the target stimuli. In this case, the stimulus is used as the main source of information to signal the response without further elaboration in the context of the sequence (Tubau et al., 2007). Here, performance improvement is evident even when the task is lacking a sequence and can be described as general learning effects (Abrahamse & Noordzij, 2011). By contrast, performance improvement from plan-based control is associated with sequence-specific representations, which means that response efficiencies observed in latency reductions are based on reducing stimulus reliance and instead rely on responses and feedback (Abrahamse & Noordzij, 2011; Robertson, 2007; Willingham, 1999). The efficiencies developed are specific to an internalised structure of the practiced sequence, and so performance gains established by plan-based control are lost when there is deviance from the learnt sequence structure.
It is important to note that although stimulus- and plan-based control modes support learning, the learner is not exclusively utilising one control mode over the other during learning. For example, both control modes can simultaneously support the reduction of reaction time in sequence learning, but when no sequence regularity is evident, then the control system relies more on perceptual processes to support learning (Robertson, 2007). Recently, it has been noted that this switch of control modes can occur dynamically and almost instantaneously in learners which indicates that control changes are not just expected between learning blocks but also within block dynamics (Verwey et al., 2015). This switching of control modes mainly serves the purpose of maximising performance efficiency during sequence learning (Abrahamse, Braem, Notebaert, & Verguts, 2016), to which different frontal regions of the cortex may play a role in regulating cognitive contributions for performance (Koechlin & Summerfield, 2007).
One of the main determinants for the switching of stimulus- or plan-based control, is the manner by which cognitive control facilitates attention towards task-relevant stimuli and/or inhibition of irrelevant information (Friedman & Miyake, 2004; Gallant, 2016). Indeed, it is posited that cognitive control is contextually driven and works dynamically in a double-edged manner (Amer, Campbell, & Hasher, 2016; Amer & Hasher, 2014). When top-down cognitive control is enhanced, a resultant convergent control style prioritises speed and accuracy for responding to the stimulus (Colzato, Ozturk, & Hommel, 2012). In contrast, a weakened top-down control facilitates a divergent control style that searches for different plans to the problem (Colzato et al., 2012) through the exploration of stimulus-stimulus or response-response associations in the SRTT (Abrahamse et al., 2010; Verwey et al., 2015). Several factors such as individual differences and age can affect the modulation of top down cognitive control. For example, Biss, Ngo, Hasher, Campbell, and Rowe (2013) compared the performance of a task that required remembering a list of words between a group of older and younger adults and used part of the words disguised as distractors during presentation. It was found that older adults rarely forgot words that were presented as distractors while younger adults forgot words in both the original list and as distractors. Younger adults utilised an enhanced top-down control approach to focus attention and suppress distractors, while older adults were more creatively using the so-called “irrelevant distractors” for rehearsal (Amer et al., 2016; Amer & Hasher, 2014; Biss et al., 2013). More specifically, enhanced/weakened top-down cognitive control is considered to support two different systems in the form of information processing and storage, and problem solving during sequence learning. Specifically, it was found that early stages of sequence learning were not age-dependent and meditated by problem solving (weakening of top-down) while early and late sequence learning are mediated by more basic cognitive control functions such as processing speed, attention and working memory (enhanced top-down) (Krüger, Hinder, Puri, & Summers, 2017). More recent work has found that cognitive tasks like meditation is able to bias cognitive control and attention in either convergent (enhanced top-down) or divergent (weakened top-down) control styles (Colzato et al., 2012; Colzato, Szapora, Lippelt, & Hommel, 2017), which in turn can affect whether stimulus- or plan-based control is prioritised when it precedes learning (Chan et al., 2017; Immink et al., 2017).
Recent investigations of single-session FAM practice supported that cognitive control was biased in a convergent control style during performance of subsequent goal-directed cognitive tasks (Colzato, van der Wel, Sellaro, & Hommel, 2016; Lippelt, Hommel, & Colzato, 2014; Lutz, Jha, Dunne, & Saron, 2015; van Leeuwen, Singer, & Melloni, 2012). For direct applications in sequence learning, when FAM immediately preceded learning, stimulus-based control was responsible for reduction in response times and improvement in general learning performance, attributed to an enhanced top-down and convergent control style (Chan et al., 2017). It was also shown that when FAM was temporally more distant relative to sequence learning, plan-based learning was responsible for greater levels of improvements in sequence-specific learning, attributed to a weakened top-down control and divergent control style (Chan et al., 2017). As it stands, it may be too simplistic to assume that the immediate effects of FAM in naïve participants operate in one mechanistic fashion of biasing convergent cognitive control in sequence learning, and that factors such as the temporal placement of FAM relative to learning, may also alter effects in cognitive control.
Another factor that may be modulating cognitive control from single-session FAM could be related with perceived effort during practice. Effort is considered a representation of the degree of cognitive demand (Fairclough & Houston, 2004; Hockey, 1997) and the mobilisation of cognitive resources needed in a task (Gendolla & Wright, 2009), which therefore may be representative of cognitive flexibility processes that can be altered from meditation practice (Malinowski, 2013). Immink et al. (2017) found that perceived effort following single-session practice of meditation exerted different effects for the cognitive control of subsequent sequence learning in naïve participants. Specifically, it was found that low effort experienced during meditation practice (regardless of type) facilitated greater levels of general learning effects. It is therefore possible that cognitive effort during FAM maybe modulating cognitive control to bias either convergent or divergent control that leads to more persistent stimulus- or plan-based learning control modes. Further cognitive control effects of effort are discussed in a later section.
Attention is featured prominently in theoretical models of meditation (Malinowski, 2013; Tang et al., 2015) as a central construct by which the cognitive system is augmented to influence mental states. One of the main issues in meditation research is the design of suitable control conditions that mimics the same cyclical phenomenon of attention in cognitive control (i.e. monitoring, disengaging, shifting, sustaining, distraction, see Malinowski, 2013) during meditation practice for operational equivalence (Davidson & Kaszniak, 2015; Hasenkamp, Wilson-Mendenhall, Duncan, & Barsalou, 2012). Attention operates to prioritise goal-relevant information and can function like a “zoom lens” (Eriksen & St. James, 1986; Lutz et al., 2015), which focus on different details during learning. This phenomenon is coined attentional aperture and describes that attentional control with an ability to adopt a narrow (tunnelled), mid or broad (attending to many global features) range in a continuum based on the engaged task (Derakshan, Ansari, Hansard, Shoker, & Eysenck, 2009; Eriksen & St. James, 1986; Ingram, 2009; Lutz et al., 2015).
Emerging research supports that the attentional aperture responds to cognitive control over global tasks and can be augmented by meditation practice in a neurobehavioural paradigm (Lutz et al., 2015). A single-session FAM has been shown to bias attentional allocation in subsequent cognitive tasks (Colzato et al., 2012; Colzato, Sellaro, Samara, Baas, et al., 2015; Colzato, Sellaro, Samara, & Hommel, 2015; Colzato et al., 2017). Specifically, a single-session of FAM in meditation naïve participants was shown to result in larger attentional blink response in the global-to-local task, consistent with the notion of an increased attentional selection for target stimuli, indicating enhanced top-down control and a biased convergent processing style (Colzato et al., 2016; Colzato et al., 2017; Lippelt et al., 2014).
Since FAM is considered to influence sequence learning through attention control, then other cognitive tasks that involve attention focusing (via narrowing effects) may be expected to exert similar result for stimulus-based planning in sequence learning (Tang & Posner, 2009). Particularly, laboratory-based computerised attentional tasks that are goal-directed, require participants to monitor stimuli through utilisation of attention and engage cognitive control to successfully meet task objectives (Basner & Dinges, 2011). When participants are less engaged or distracted, lapses in attention can occur, which are in a similar vein to cyclical cognitive control processes during meditation practice, where reengagement of task requires disengagement of the distraction to maintain focus. Computerised attentional tasks are therefore operationally targeted towards the utilisation and engagement of attentional control.
Although attention control in FAM is centralised, it is also a cognitive resource shared between other core regulatory processes such as cognitive flexibility and emotion regulation, thought to be influenced by meditation practice (Malinowski, 2013; Tang et al., 2015). Thus, it is possible that improved general learning and sequence-specific learning effects following FAM (Chan et al., 2017) might not only be based on the enhancement of cognitive control and attentional processes, but also due to other moderators of motor learning performance such as cognitive effort, arousal and pleasure. Earlier, we presented work that meditation increased mobilisation of cognitive resources resulting in subsequent enhanced motor performance (Immink et al., 2017). If meditation enhances sequence learning by increasing cognitive effort, it is then reasonable to suggest that similar preceding attention task might also increase effort resulting in sequence learning enhancement. Indeed, Borragan, Slama, Destrebecqz, and Peigneux (2016) showed that prior engagement in an unrelated computerised task that required high cognitive effort, improved sequence-specific learning. The improvement in sequence-specific learning was attributed to cognitive fatigue effects, which subsequently facilitated a weakened top-down cognitive control state to increase plan-based control. This evidence together with the results of Immink et al. (2017) show that participation in different cognitive tasks may bias cognitive control differently and therefore facilitate the adoption of either stimulus- and plan-based control strategies dependent on the level of engagement (Amer et al., 2016).
In addition to effort effects, FAM practice is also commonly associated with the inducement of positive pleasure states (Hamilton, Kitzman, & Guyotte, 2006), low in arousal, and a relaxed and tranquil but alert state of mind (Kabat-Zinn et al., 1992; Lazar et al., 2000; Warrenburg, Pagano, Woods, & Hlastala, 1980). FAM might therefore also influence cognitive control in sequence learning through other alternative mechanisms such as arousal (Malinowski, 2013) and pleasure states (Hommel, 2015; van Steenbergen, Band, & Hommel, 2010), although there appears to be a lack of direct evidence of these effects from FAM on cognitive control. In essence, low arousal and positive pleasure states are typically associated with simple heuristics associated with reduced cognitive effort, more creative approaches, and a weakened top-down cognitive control (Schwarz, 1990), biased towards more divergent cognitive control styles (Hommel, 2015). Indeed, Shang, Fu, Dienes, Shao, and Fu (2013) showed that low arousal and increased pleasure states (through music) benefited sequence learning through increased plan-based control, although no differences in reaction time latency reductions for the SRTT between positive and negative states were found.
Conversely, negative or reduced pleasure states may have narrowing effects on attention and can occur independent of arousal changes (Easterbrook, 1959). It is thought that the narrowing acts as a shield from distraction by increasing effort and perceptual focus, thus enabling concentration and focus on a narrow stream of information at the expense of excluding relevant information (van Steenbergen et al., 2010). This process is also suggestive of enhanced top-down cognitive control that may facilitate stimulus-based control. Pretz, Totz, and Kaufman (2010) found support for negative pleasure states to enhanced general learning performance in an artificial grammar sequencing task, and no benefit of positive mood in enhancing general or sequence-specific learning performance. It is also possible that single-session FAM practice in naïve participants could induce negative pleasure states due to feelings of unfamiliarity of practice or simply a general dislike for meditation. Such feelings may therefore induce reduced pleasure states and enhance cognitive control to prioritise stimuli processing due to attentional narrowing. Simply put, the effects from meditation on core regulatory processes are poorly understood and it is important to establish their links to cognitive control. Effort, arousal and pleasure states can be manipulated by FAM and it is important to understand whether these effects have direct, indirect, or operate in parallel to influence cognitive control for sequence learning performance.
The goal of the current experiment was to test the effect of a single-session of FAM on subsequent motor sequence learning and whether FAM operated on a framework based on attentional control as a centralised mechanism for single-session effects. We proposed to compare FAM against a computerised attention task (CAT) as an active control task to strengthen our study design, in addition to a rest-only control group. The current experiment design also allowed for the replication of testing an earlier finding by Chan et al. (2017) that a single-session FAM could immediately bias cognitive control to influence subsequent sequence learning for general learning performance improvements against a rest-only control group. We predict that exposure to a single-session of FAM will enhance top-down cognitive control in naïve participants by narrowing attentional control such that subsequent sequence learning relies primarily on stimulus-based control under the notion of general learning (Abrahamse & Noordzij, 2011), and not because of plan-based control under the notion of sequence-specific learning (Robertson, 2007; Willingham, 1999; Willingham, Wells, Farrell, & Stemwedel, 2000).
We also considered whether meditation related perceived effort, arousal and pleasure, would provide evidence to explain the immediate effects on sequential learning performance. We predict that if FAM related effort is high, arousal high, and pleasure low, a more convergent cognitive control style would be resultant and therefore allow for benefits in general practice effect where decreasing response latencies are not attributable to a sequence structure. In contrast, if FAM related effort is low, arousal low, and pleasure high, a more divergent cognitive control style would be resultant and therefore allow for benefits in sequence-specific learning where improvements in response latencies are only evident when the sequence structure established during practice is available.
Section snippets
Participants
Fifty-five meditation naïve volunteers were recruited from a University course to participate in the present experiment in exchange for partial credit towards the course. We applied experimental exclusion criteria where participants needed to have at least 6 h of sleep, no more than one standard alcoholic drink the night before, must not practice any form of meditation or taken part in a motor learning study within the last year by checking their names against our experimental database. Of
Familiarisation performance
LMER model analysis of familiarisation performance revealed that there was neither a significant Group × Response error interaction for trial RT (p = .62), nor a significant Group × Response error × Block interaction for trial RT (p = .07). Thus, Response error across groups were similar (FAM:2.1%, CAT:2.8%, Control:2.6%) and the model was simplified by removing all incorrect trials and Response error as a factor for analysis of RT performance. This resulted in the exclusion of 2.4% of all SRTT
Discussion
Previously, Chan et al. (2017) demonstrated that a single-session of FAM enhanced subsequent sequential behaviour performance. Enhanced performance appeared to be afforded by efficient implementation of stimulus-based control as opposed to plan-based control strategies. This was interpreted to indicate that FAM states enhanced top-down control, which are then implemented in subsequent sequence learning. As FAM is thought to involve the practice of attention control processes (Malinowski, 2013;
Acknowledgements
We thank Mr. David Burgess for providing a recording of the yoga nidra meditation used in this experiment and Dr. Phillip M. Alday for early advice on statistical modelling.
References (77)
- et al.
Designing training programs for perceptual-motor skills: Practical implications from the serial reaction time task
Revue Européenne de Psychologie Appliquée
(2011) - et al.
Cognitive control as a double-edged sword
Trends in Cognitive Sciences
(2016) - et al.
The influence of focused-attention meditation states on the cognitive control of sequence learning
Consciousness and Cognition
(2017) - et al.
Meditation-induced states predict attentional control over time
Consciousness and Cognition
(2015) - et al.
Meditation-induced cognitive-control states regulate response-conflict adaptation: Evidence from trial-to-trial adjustments in the Simon task
Consciousness and Cognition
(2015) - et al.
A single bout of meditation biases cognitive control but not attentional focusing: Evidence from the global-local task
Consciousness and Cognition
(2016) - et al.
A metabolic measure of mental effort
Biological Psychology
(2004) Mindfulness meditation practice and executive functioning: Breaking down the benefit
Consciousness and Cognition
(2016)- et al.
Mind wandering and attention during focused meditation: A fine-grained temporal analysis of fluctuating cognitive states
NeuroImage
(2012) Between persistence and flexibility: The yin and yang of action control
Better control with less effort: The advantage of using focused-breathing strategy over focused-distraction strategy on thought suppression
Consciousness and Cognition
An information theoretical approach to prefrontal executive function
Trends in Cognitive Sciences
Is meditation always relaxing? Investigating heart rate, heart rate variability, experienced effort and likeability during training of three types of meditation
International Journal of Psychophysiology
Attention regulation and monitoring in meditation
Trends in Cognitive Sciences
The unity and diversity of executive functions and their contributions to complex “Frontal Lobe” tasks: A latent variable analysis
Cognitive Psychology
The psychology experiment building language (PEBL) and PEBL test battery
Journal of Neuroscience Methods
Attentional requirements of learning: Evidence from performance measures
Cognitive Psychology
A formal model of capacity limits in working memory
Journal of Memory and Language
The effects of mood, cognitive style, and cognitive ability on implicit learning
Learning and Individual Differences
Attention training and attention state training
Trends in Cognitive Sciences
Training brain networks and states
Trends in Cognitive Sciences
Learning a keying sequence you never executed: Evidence for independent associative and motor chunk learning
Acta Psychologica
Grounding cognitive control in associative learning
Psychological Bulletin
Representing serial action and perception
Psychonomic Bulletin & Review
Conceptual processing of distractors by older but not younger adults
Psychological Science
The influence of Buddhist meditation traditions on the autonomic system and attention
BioMed Research International
Maximizing sensitivity of the psychomotor vigilance test (PVT) to sleep loss
Sleep
Fitting linear mixed-effects models using lme4
Distraction can reduce age-related forgetting
Psychological Science
Cognitive fatigue facilitates procedural sequence learning
Frontiers in Human Neuroscience
Toward understanding the relationship between feeling states and social behavior
Meditate to create: The impact of focused-attention and open-monitoring training on convergent and divergent thinking
Frontiers in Psychology
Prior meditation practice modulates performance and strategy use in convergent- and divergent-thinking problems
Mindfulness
Neurocognitive mechanisms of statistical-sequential learning: What do event-related potentials tell us?
Frontiers in Human Neuroscience
Conceptual and methodological issues in research on mindfulness and meditation
The American Psychologist
Anxiety, inhibition, efficiency, and effectiveness. An investigation using antisaccade task
Experimental Psychology
The effect of emotion on cue utilization and the organization of behavior
Psychological Review
Visual attention within and around the field of focal attention: A zoom lens model
Perception & Psychophysics
Cited by (10)
Focused-attention meditation increases cognitive control during motor sequence performance: Evidence from the N2 cortical evoked potential
2020, Behavioural Brain ResearchCitation Excerpt :Increased reliance on plan-based control suggests that cognitive control is weakened in the delay between FAM and sequence learning [7]. Follow up work provided further evidence that when a single-session of FAM immediately precedes sequence learning, the pattern of learning reflects heightened general learning as opposed to sequence-specific learning [3]. In this work, a session of FAM was compared to completion of a visual attention task, which was equal in duration.
The Influence of Focused Attention and Open Monitoring Mindfulness Meditation States on True and False Memory
2023, Journal of Cognitive EnhancementA Single Session of Mindfulness Meditation Expedites Immediate Motor Memory Consolidation to Improve Wakeful Offline Learning
2023, Journal of Motor Learning and DevelopmentInstantaneous effects of mindfulness meditation on tennis return performance in elite junior athletes completing an implicitly sequenced serve return task
2022, Frontiers in Sports and Active Living
- ☆
This research was supported by the Australian Government Research Training Program Scholarship for the first author.