Periodic Reporting for period 1 - Time in Action (Shaping time through action learning)
Período documentado: 2021-09-01 hasta 2023-08-31
Timing is deeply rooted in human cognition and behaviour, providing the basis for our ability of comprehending speech, playing sports, and appreciating music. Despite lack of dedicated sensory and neural systems for time, people exhibit extraordinarily precise time estimation. Our sense of time is believed to leverage motor systems endowed with high temporal precision required for action control. This relationship is reflected in a large body of behavioural studies showing that time estimation is significantly distorted by action. However, this literature is represented by experimental approaches that overlook critical aspects of motor function. Actions are goal driven and control is gradually refined through learning, yet our understanding of how the motor system instantiates time stems from paradigms with ill-defined behavioural goals, and where timing operation is captured through static snapshots of average performance, not accounting for dynamic aspects of action learning, and the way these processes gradually shape timing.
This fundamental gap in the literature has motivated this Marie Curie IF project (Time in Action), aimed at understanding how timing is shaped in the visual domain during action planning, and critically how this relationship is modulated by goal driven motor control & motor learning processes. I have set up 3 experiments to meet these aims, combining perceptual judgements of duration, hand tracking and electroencephalographic measurements of brain activity (EEG), to understand how the brain constructs an understanding of time through action control, and what neural mechanisms underlie this process. Time in Action provides a paradigm shift in this literature by accounting for goal-driven action learning mechanisms that are central the motor system’s operation, and the way these processes dynamically shape our sense of time.
This fundamental gap in the literature has motivated this Marie Curie IF project (Time in Action), aimed at understanding how timing is shaped in the visual domain during action planning, and critically how this relationship is modulated by goal driven motor control & motor learning processes. I have set up 3 experiments to meet these aims, combining perceptual judgements of duration, hand tracking and electroencephalographic measurements of brain activity (EEG), to understand how the brain constructs an understanding of time through action control, and what neural mechanisms underlie this process. Time in Action provides a paradigm shift in this literature by accounting for goal-driven action learning mechanisms that are central the motor system’s operation, and the way these processes dynamically shape our sense of time.
Time in Action involved a dual hand reaching / time estimation experimental setup that provided a behavioural testbed to study this action learning / time perception relationship (Fig 1). Participants timed brief visual stimuli (probes) while 1) planning fast (ballistic) index reaching movements aimed at cued goal locations (Planning), 2) during the period preceding unplanned movements aimed at uncued goals (Unplanned - Motor control), and 3) during matched intervals involving no reaching movements (No Reach - Sensory control) (Fig 2).
I studied how these estimates vary throughout the course of the experiment, as a function of action learning. These conditions were tested in two experiments in which I independently manipulated types of action performed by groups of 20 participants (directional: left or right, or distance: near or far). These experiments revealed that time expands when planning action (opposed to unplanned actions or sensory control), and this expansion becomes progressively stronger throughout the course of the experiment. Computational models of learning revealed that time expands as they acquire more experience at the motor task, reflected in increases of motor readiness (how quickly they initiate a planned action; Fig 2). EEG responses were analyzed to identify the neural source of these effects. Analyses show that action and timing processes interact across visual and motor control brain areas (Fig 1d), where motor planning operations (1) shape computations in early visual (2) and higher order timing stages (3).
These findings reveal that our sense of is time ‘rides’ on predictions we make about our interactions with the environment, and this relationship is strengthened as actions are optimized through experience.
But how can we properly time our behaviours if actions need to rely on distorted perceptual estimates? This seems at odds with the fact that humans are clearly able of properly timing behaviours in response to timed sensory events, such as correctly estimating traffic light durations to time when to place our foot on the gas. Experiment 3 was aimed at answering this question, where participants simultaneously judge the duration of a stimulus during action planning (sensory timing - as in Experiments 1 and 2), while relying on the same stimulus to time a motor response (sensorimotor timing). This experiment shows that our conscious experience of duration and our motor system’s ability of timing sensory events to guide action, are partially decoupled processes. While we might experience a visual stimulus as being long as we plan behaviours, our motor system is capable of estimating the same stimulus correctly when planning action.
Two scientific papers are in preparation detailing these findings, that I plan to submit in early 2024.
I studied how these estimates vary throughout the course of the experiment, as a function of action learning. These conditions were tested in two experiments in which I independently manipulated types of action performed by groups of 20 participants (directional: left or right, or distance: near or far). These experiments revealed that time expands when planning action (opposed to unplanned actions or sensory control), and this expansion becomes progressively stronger throughout the course of the experiment. Computational models of learning revealed that time expands as they acquire more experience at the motor task, reflected in increases of motor readiness (how quickly they initiate a planned action; Fig 2). EEG responses were analyzed to identify the neural source of these effects. Analyses show that action and timing processes interact across visual and motor control brain areas (Fig 1d), where motor planning operations (1) shape computations in early visual (2) and higher order timing stages (3).
These findings reveal that our sense of is time ‘rides’ on predictions we make about our interactions with the environment, and this relationship is strengthened as actions are optimized through experience.
But how can we properly time our behaviours if actions need to rely on distorted perceptual estimates? This seems at odds with the fact that humans are clearly able of properly timing behaviours in response to timed sensory events, such as correctly estimating traffic light durations to time when to place our foot on the gas. Experiment 3 was aimed at answering this question, where participants simultaneously judge the duration of a stimulus during action planning (sensory timing - as in Experiments 1 and 2), while relying on the same stimulus to time a motor response (sensorimotor timing). This experiment shows that our conscious experience of duration and our motor system’s ability of timing sensory events to guide action, are partially decoupled processes. While we might experience a visual stimulus as being long as we plan behaviours, our motor system is capable of estimating the same stimulus correctly when planning action.
Two scientific papers are in preparation detailing these findings, that I plan to submit in early 2024.
Time in Action provides a significant stride forward in our understanding of time perception, and the way it is integrated with our ability of planning behaviour and successfully interacting with our environment. Time in Action provides a new methodological and conceptual framework for investigating and understanding how time estimation is shaped and constantly updated based on the operation of the motor system in a learning based behavioural paradigm. The findings add an important piece to the time perception puzzle by showing that our sense of time is intimately coupled with our ability of planning behaviour and anticipating its consequences, and this relationship is strengthened as actions are optimized through experience, thus highlighting the dynamic interplay between action control, action learning and ongoing estimates of sensory duration. These data and methodologies also have significant implications in the clinical understanding and diagnosis of conditions that jointly affect motor control and time perception, such as Parkinson’s Disease, Huntington’s Disease and Tourette’s syndrome. This is methodologically afforded by a paradigm that is designed to jointly probe dynamic motor control/learning and time estimation, which can be repurposed for clinical study contexts, and the fact that I provide data analysis approaches to relate the evolution of these temporal and motoric ongoing processes.