Spatial construal of time and experimental design.

(A) The schematic diagram of spatial construal of time. It illustrates two core time concepts (sequence and duration) and two major perspectives on event series (mental time travel and watching). (B) Stimulus: a fictional religious ritual of 15 events following a specific sequence, enduring particular durations, and happening on predetermined parts of the day. To minimize potential confounds between the semantic content of the event phrases and the temporal structure of the events, we randomly assigned the phrases to the events, creating two versions for participants with even and odd ID numbers. Only one version is illustrated here. Both versions can be seen in Figure1—figure supplement 1 and Figure 1—source data 1. (C) Task paradigm. In the external-perspective task, participants judged whether the target events happened in the same part of the day as the reference event. In the internal-perspective task, participants imagined themselves doing the reference events and judged whether the target event happened in the past or will happen in the future.

Reaction time analysis.

For diagnostic purposes, we plotted the partial residuals of each significant predictor significantly influencing the reaction time (RT) of the corrected trials. The partial residual includes the effect of each variable, its interaction with Task Type, and the residuals from the full linear regression model. A. RT increased with Syllable Length, showing a similar trend across both tasks. B. Sequential Distance affected RT in opposite directions depending on the perspectives. C. Event Duration influenced RT only in the external-perspective task, with no effect in the internal-perspective task.

The reaction time of the corrected trials indicates that time is differently processed under internal and external perspectives1.

1Linear Mixed Model Formula: RT ∼ 1 + Task Type * (Sequential Distance + Duration + Syllable Length) + (1 | Participant) 2The significant effects were highlighted. We did not highlight the significant main effects if the corresponding interaction effects were also significant.

Neural correlates of specific perspectives and syllable length.

(A) Univariate contrast between external-perspective and internal-perspective tasks (voxel-level p < 0.001, cluster-level FWE corrected p < 0.05). All the significant areas were in the right hemisphere. PreC: precuneus; RSC: the retrosplenial cortex; SFG: the superior frontal gyrus; AG: angular gyrus; SMA: supplementary motor area; SMG: supramarginal gyrus. (B) Parametric modulation of syllable length as a sanity check (voxel-level p < 0.001, cluster-level FWE corrected p < 0.05). The activation level in the anterior part of the left superior temporal gyrus and the visual cortex positively correlated with syllable length. R = Right Hemisphere; L = Left Hemisphere.

Univariate contrast between internal- and external-perspective tasks (p < 0.001, cluster-level FWE corrected p < 0.05 across the whole cortex).

1The average Montreal Neurological Institute coordinates of all the significant voxels of each cluster. The precuneus and the retrosplenial cortex were connected as one cluster under the threshold p < 0.001 (z > 3.09). In this case, we increased the threshold to the point when the precuneus and the retrosplenial cortex were separate (z > 3.3) and calculated the average coordinates of each cluster. 2The voxel size is 3 × 3 × 3 mm3.

Neural correlates of event sequence.

A-C: Interaction effect between Task Type (i.e., external-vs. internal-perspective tasks) and Sequential Distance. (A) The only cortical region showing a significant interaction effect was localized in the right posterior parietal cortex (voxel-level p < 0.001, cluster-level FWE corrected p < 0.05). (B) Regions of interest analysis shows that the activation level in the right posterior parietal cortex correlated with sequential distance positively in the external-perspective task and negatively in the internal-perspective task. (C) A further illustration of the relations between the activation level in the right posterior parietal cortex and sequential distance in the two tasks. The error bar indicates the standard error relative to the mean, and the shaded band around the linear regression line indicates 95% confidence interval. D-F: main effect of Sequential Distance. (D) The right hippocampal head shows a significant main effect of Sequential Distance within the mask of the bilateral hippocampus (voxel-level p < 0.001, cluster-level FWE corrected p < 0.05; voxel-level p < 0.05 for illustration purposes). (E) Regions of interest analysis shows that the correlation between the activation level in the right hippocampal head and the sequential distance was independent of perspectives. (F) A further illustration of the relations between the activation level in the right hippocampal head and sequential distance in the two tasks. The error bar indicates the standard error relative to the mean, and the shaded band around the linear regression line indicates 95% confidence interval. R = Right Hemisphere. **: p < 0.01; ***: p < 0.001.

Neural correlates of event duration.

(A) The right hippocampal body shows a significant main effect of Duration within the mask of the bilateral hippocampus (voxel-level p < 0.001, cluster-level FWE corrected p < 0.05; voxel-level p < 0.05 for illustration purposes). (B) However, regions of interest analysis shows that the correlation between the activation level in the right hippocampal body and Duration significantly differs in the internal- and the external-perspective task. (C) A further illustration of the relations between the activation level in the right hippocampal body and duration in the two tasks. The error bar indicates the standard error relative to the mean, and the shaded band around the linear regression line indicates 95% confidence interval. (D) Directly comparing the effects of Sequential Distance and Duration in the head and the body of the hippocampus shows a double dissociation pattern: the hippocampal head represented Sequential Distance but not Duration, while the hippocampal body represented Duration but not Sequential Distance. R = Right Hemisphere. ***: p < 0.001; n.s.: not significant.

All participants learnt the same event structure.

However, the event order differed across participants (i.e., two event lists were used for even- and odd-numbered participants). This approach helps minimize potential confounding between the temporal and semantic information carried by the events.

Reaction time analysis including Parts of the Day.

For diagnostic purposes, we plotted the partial residuals of each significant predictor significantly influencing the reaction time (RT) of the corrected trials. The partial residual includes the effect of each variable, its interaction with Task Type, and the residuals from the full linear regression model. A. RT increased with Syllable Length, showing a similar trend across both tasks. B. Sequential Distance affected RT in opposite directions depending on perspectives. C. Event Duration influenced RT only in the external-perspective task, with no effect in the internal-perspective task. D. The effect of Parts of Day was in line with the effect of Sequential Distance, with two events in the same and different parts of the day corresponding to the short and long sequential distances, respectively.

(A) Brain surface view of the univariate contrast between external-perspective and internal-perspective tasks (voxel-level p < 0.001, cluster-level FWE corrected p < 0.05). This view is transformed from the significant clusters in the MNI space in Figure 3A to the fsLR space using the toolbox neuromaps (https://github.com/netneurolab/neuromaps). All the significant areas were in the right hemisphere. (B) Brain surface view of the default network A and the default network B. They were respectively identified as the 8th and the 1st components among the 25 components of the “group-ICA” template from the UK Biobank brain imaging (https://www.fmrib.ox.ac.uk/ukbiobank/). We preserved the positive voxels above 7 in the MNI space and transformed them into the fsLR space using the toolbox neuromaps. Both plots are illustrated using the Connectome Workbench 2.0 (https://www.humanconnectome.org/software/connectome-workbench). They are displayed on an inflated surface against the group-averaged all sulcus image of 1096 young adults from the dataset of the Human Connectome Project (https://balsa.wustl.edu/reference/pkXDZ).