Task errors drive memories that improve sensorimotor adaptation

Abstract

11Traditional views of sensorimotor adaptation, or adaptation of movements to 12 perturbed sensory feedback, emphasize the role of automatic, implicit correction of 13 sensory prediction errors (differences between predicted and actual sensory 14 outcomes). However, latent memories formed from sensorimotor adaptation, 15 prominently evidenced in improved learning (i.e., savings), have recently been 16 attributed to strategic corrections of task errors (failures to achieve task goals). To 17 dissociate contributions of task errors and sensory prediction errors to latent 18 sensorimotor memories, we perturbed target locations to remove or enforce task 19 errors during learning and/or test. We show that prior learning to correct task errors 20 was sufficient for savings: a history of sensory prediction errors was neither sufficient 21 nor obligatory for savings. Limiting movement preparation time further revealed two 22 distinct components of this learning: 1) time-consuming, flexible strategies, and 2) 23 rapidly expressible, inflexible stimulus-response associations. The results show that 24 adaptive responses to sensorimotor perturbations take many forms. Herzfeld et al., 2014). According to this idea, experiencing a 66 systematic sequence of errors generates a memory of errors, which increases 67 sensitivity to those errors and the gain of error correction when reencountering 68 similar errors (Gonzalez Castro et al., 2014; Herzfeld et al., 2014). 65 Castro 69Despite the importance of the phenomenon of savings for the broad field of 70 3 motor learning, interest in the processes that underlie the effect has declined since 71 the emergence of compelling evidence that savings in standard visuomotor rotation 72 tasks is dominated by volitional, strategic selection of actions that restore task 73 success (Avraham et al., 2019; de Brouwer et al., 2017; Haith et al., 2015; Morehead 74 et al., 2015). For example, if leftward displacement of visual feedback causes a 75 person to miss the target to the left, they tend to deliberately aim to the right of the 76 target to counteract the error (Uhlarik, 1973). Using cues to signal perturbation onset 77 prompted rapid re-selection of the adapted movement on the first trial, even before 78 participants re-experienced the previously encountered error (Morehead et al., 79 2015). Similarly, suppressing strategy use by limiting movement preparation time 80 prevented the expression of savings (Haith et al., 2015). If savings were solely a 81 manifestation of deliberate strategy use, then "cognitive" functions which determine 82 how successfully one might acquire and implement strategies would be central to 83 acquisition of latent sensorimotor memories: non-strategic "motor" processes would 84 thus be of less importance. 85 But is the conscious awareness of a strategy, to apply a corrective movement 86 that already exists within the participant's motor repertoire, really all that is retained 87 when a person adapts movement to a perturbation? Or are there components of ...

Acoustic stimulation increases implicit adaptation in sensorimotor adaptation

“…We found that LAS boosted adaptation to sensory prediction errors under no-task-error conditions, particularly at initial exposure to the perturbation. Under standard task error conditions, this effect of LAS might have been masked by task-error driven components of adaptation (such as explicit re-aiming strategies (Mazzoni & Krakauer, 2006) or stimulus-response associations) (Ishii et al, 2018;McDougle & Taylor, 2019;Leow et al, 2020), as it was not discernible under standard task error conditions. In Experiment 2, we further explored the capacity of LAS to influence the retention of sensorimotor memories.…”

“…Participants were assigned to the Standard Task Error LAS group (n=16), Standard Task Error no LAS group (n =16), the No Task Error LAS group (n=16) or the No Task Error no LAS group (n =16). Sample size selection was based on similar work employing task error manipulations (Leow et al, 2020).…”

“…In target-reaching tasks, task errors often take the form of discrepancies between the target position and movement endpoint (Schaefer et al, 2012;Gaveau et al, 2014;Reichenthal et al, 2016), although they can also take the form of failures to achieve more abstract performance goals (Mazzoni & Krakauer, 2006;Taylor & Ivry, 2011). Task errors can elicit the use of explicit strategies (such as volitionally re-aiming to the left of a target to counteract a clockwise rotation of cursor feedback) (Uhlarik, 1973;Mazzoni & Krakauer, 2006), and the formation of associations between cues or stimuli properties within the task context and behavioural responses (Welch, 1971;Cunningham & Welch, 1994;Ishii et al, 2018;McDougle & Taylor, 2019;Leow et al, 2020).…”

Hierarchical motor adaptations negotiate failures during force field learning

“…On the other hand, recent studies have shown that the TE modulates the adaptation rate of the SPE-driven internal model adaptation (Kim et al, 2019) or savings (Leow et al, 2020). This role of the TE as a modulator to the internal model adaptation may suggest a hierarchical interaction rather than parallel interaction between the TE-driven and the SPE-driven motor adaptations.…”

“…The interaction between the explicit strategy learning and the internal model adaptation is popularly explained by a two-state model of sensorimotor learning (Smith et al, 2006), where the two operate in parallel and the net adaptation is defined to be the sum of the two (McDougle et al, 2015, Miyamoto et al, 2020). On the other hand, recent studies have shown that the TE modulates the adaptation rate of the SPE-driven internal model adaptation (Kim et al, 2019) or savings (Leow et al, 2020). This role of the TE as a modulator to the internal model adaptation may suggest a hierarchical interaction rather than parallel interaction between the TE-driven and the SPE-driven motor adaptations.…”