Your Sense of Time Stretches as You Learn to Reach Faster

Time, as your brain experiences it, is not a fixed ruler laid over the world. It is a moving target that bends around what you are about to do. A new study now shows that this bending is not a one-off quirk of action preparation but something that evolves with learning: as people practiced rapid hand-reaching movements over many trials, their perception of brief visual durations stretched progressively, and the amount of stretching tracked how much faster they became at initiating each reach. The findings suggest that action planning acts as an internal metric for time estimation, one that is recalibrated trial by trial as the motor system refines its control.

The research, published in iScience, combined behavioral measurement, electroencephalography (EEG), and computational modeling to ask how conscious timing judgments are shaped throughout goal-directed action control. The question matters because timing is pivotal to cognition and behavior. Humans estimate durations in the millisecond-to-second range with striking precision, despite the absence of any single sensory or neural system uniquely dedicated to timing. Converging evidence suggests the brain sidesteps this constraint by exploiting motor control circuitry, which is endowed with the millisecond-level temporal precision required for scheduling muscle activations. At its core, action control is a temporal problem: everything from walking to striking a match-winning penalty kick hinges on precisely timed commands.

Neural and behavioral evidence supports this motoric foundation of time. Brain studies point to a spatially distributed timing architecture rooted in a core timing network comprising the supplementary motor area, the basal ganglia, and the cerebellum, augmented by auxiliary areas flexibly recruited according to task demands. Behavioral studies reveal significant overlap between timing and action: perceptual time is distorted in proximity of behavior. Time expands as we prepare actions, compresses while we carry them out, and expands again as we evaluate behavioral consequences. These effects are modulated by movement direction, force, frequency, and speed, generally reflecting a compatibility between action and perceptual magnitudes, with stronger, faster, and larger movements expanding perceived time.

Yet most previous action-timing studies overlooked two critical aspects of motor function: that actions are frequently goal driven, and that action control is gradually refined through practice. Tasks with little constraint on movement parameters, or with simple actions operating at ceiling-level performance, provide no headroom for improvement and no way to track changes in action performance across time. The new study addressed this gap directly. Participants performed ballistic horizontal hand reaches with their index fingers on a Plexiglass surface beneath a monitor, aiming at circular targets displayed on screen while an infrared camera tracked their movements. On-screen target locations varied by direction (left or right) in a Directional task, or by distance (near or far) in a Distance task, relative to a fixation point.

The crucial manipulation concerned what participants knew before they moved. In cued trials, the goal location was shown for one second during a variable foreperiod, allowing participants to prepare a specific reach before the acoustic Go signal. In non-cued trials, the target appeared only at the Go signal, forcing action selection immediately before movement initiation. During the foreperiod, participants judged the duration of sub-second visual test stimuli presented at central fixation, comparing them to a reference standard duration refreshed throughout the experiment. A matched sensory control condition required the same timing judgments without any motor response. At the end of each trial, participants saw where their index finger had landed, receiving accuracy feedback and a score proportional to the distance from the goal center.

The average results were clear: visual time was perceptually expanded by action planning. Timing estimates expanded prior to action onset in cued compared to non-cued actions and sensory conditions, in both the Directional and Distance tasks. When motor plans were ready to be executed at the Go signal, subjective durations grew longer. Notably, spatial features of the planned action, such as left versus right or near versus far reaches, and the mapping between hand movements and cursor motion, had no modulatory effect on timing estimates. Sensitivity to duration, measured as the just noticeable difference between test durations, did not differ across conditions either. Perceptual time, it seems, relates to the temporal aspects of action preparation, specifically when actions were initiated, rather than to non-temporal action parameters such as where movements were directed.

The trial-by-trial models revealed a more nuanced and arguably more striking relationship. Within participants, visual time progressively expanded as action latency, the time required to initiate a reach following the Go signal, decreased across trials, in both cued and non-cued conditions. Log-transformed action latencies consistently decreased over trials, while test-longer responses increased correspondingly. The reduction in latency was more pronounced in non-cued trials, likely because these offered a larger margin for optimization as participants learned to cope with target location uncertainty, perhaps by simultaneously preparing competing actions or by streamlining reactive planning triggered by the Go signal. Either way, both strategies can be honed through practice, leading to faster selection, planning, and initiation of actions.

At the participant level, the coupling became even more apparent. Individual parameters capturing each person’s rate of change in timing judgments across trials negatively correlated with their rates of change in action response latencies. Participants with greater rates of decrease in action latency exhibited greater increases in test-longer responses across trials. A sanity check excluding data from the first block of each condition yielded consistent results, suggesting the relationship reflects a task-wide process rather than transient performance fluctuations occurring early on. The study’s authors are careful to note that, given the correlational nature of the analysis, this relationship should not be interpreted as causal. Both measures may reflect common time-dependent processes accompanying performance improvements, including proceduralization of action planning, reward expectation, motivational engagement, or other practice-related factors.

The EEG data added a neural dimension to these behavioral findings. Evoked response potentials time-locked to the offset of the test stimulus showed significant effects of duration and condition, with the shortest duration eliciting a large positive potential that diminished exponentially with longer durations. This mirrors recent EEG work demonstrating distinct neural signatures of visual time peaking at frontocentral electrodes roughly 300 to 500 milliseconds after stimulus offset, reflecting the progressive accumulation of timing evidence toward a perceptual decision. Crucially, time-decoding amplitudes were modulated by action conditions: both cued and non-cued conditions elicited more negative potentials than the sensory control, with average voltage inversely related to planning specificity. Neurometric curves showed a shift in the voltage-duration relationship for cued relative to non-cued and sensory conditions, suggesting that neural activity associated with motor preparation biases concurrent time decoding.

Representational similarity analysis (RSA) then identified where and when neural activity patterns more closely resembled biased time perception relative to unbiased perception. The results showed that perceptual distortions of time were reflected in neural responses during both time encoding, as the test stimulus unfolded, and time decoding, following stimulus offset when temporal information is read out to inform the duration judgment. Encoding-related effects were selectively observed in the comparison between cued and non-cued action conditions, consistent with an association specifically with motor planning. Decoding-related effects were observed in both cued versus non-cued and cued versus sensory comparisons, suggesting a broader association with motor engagement and anticipation. In the Directional task, encoding effects appeared at central and bilateral posterior electrodes following test onset, while decoding effects emerged at frontocentral and central electrodes following test offset. The Distance task showed similar patterns that did not reach significance for decoding.

These findings align with broader evidence linking action-timing interactions to sensorimotor integration mechanisms. Previous work using magnetoencephalography showed that two types of alpha rhythm modulation associated with visual stimulation and motor preparation interact in the supplementary motor area, a key region in perceptual timing, producing a neural shift that can explain the subjective lengthening of perceived time. Importantly, time expansion in prior studies emerged only for filled visual intervals, those defined by continuous stimulation, and not for empty intervals defined only by onset and offset transients. This means the timing effect requires ongoing sensory stimulation modulated by motor preparation, further excluding non-specific attentional accounts that would affect timing irrespective of how visual information is delivered. The presence of time expansion within the non-cued condition of the new study, where the visual display was identical on every trial featuring only a central, uninformative cue, likewise rules out explanations based on visuo-attentional differences between conditions.

The broader picture emerging from this work is that perceptual time is not static but dynamically recalibrated through goal-directed action. As participants performed hand reaches with varying degrees of planning specificity, perceived duration progressively expanded, driven not by execution accuracy but by gradual decreases in the timing of action selection and initiation. Trial-wise coupling between time expansion and reduced movement latency highlights the role of motor planning temporal dynamics in shaping our perceptual sense of time. The authors position action planning as an internal metric for time estimation and establish a learning-sensitive framework linking perceptual time to the evolving dynamics of goal-driven behavior. A further limitation concerns the averaging approach applied to the EEG analyses, which characterizes overall differences in neural activity but obscures trial-wise variations that may reveal learning-related changes. Future studies combining trial-level behavioral and neural analyses may help determine how perceptual and motoric performance changes interact at the neural level.

For now, the study offers a suggestive, if still correlational, glimpse of how time perception and motor learning move together: participants whose reach initiation sped up the most across trials also showed the greatest progressive lengthening of perceived visual time. Time, in this view, may be measured against the readiness of one’s own body. The motor system, honed through learning, appears to provide the ruler against which brief visual events are judged, and as that ruler sharpens with practice, the moments before action may come to feel just a little longer.

Subject of Research: Perceptual lengthening of visual time throughout learning of hand-reaching actions

Article Title: Perceptual lengthening of visual time throughout learning of hand-reaching actions

Article References: Binetti, N., Mancinelli, F., Zanon, M., & Bueti, D. (2026). Perceptual lengthening of visual time throughout learning of hand-reaching actions. iScience, 29(11), Article 117993. https://doi.org/10.1016/j.isci.2026.117993

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117993

Keywords: time perception, action planning, motor learning, EEG, hand reaching, sensorimotor integration, duration judgment, action latency, representational similarity analysis, iScience, Perceptual, lengthening