Characteristics of motor resonance predict the pattern of flash-lag effects for biological motion

Abstract

Background: When a moving stimulus and a briefly flashed static stimulus are physically aligned in space the static stimulus is perceived as lagging behind the moving stimulus. This vastly replicated phenomenon is known as the Flash-Lag Effect (FLE). For the first time we employed biological motion as the moving stimulus, which is important for two reasons. Firstly, biological motion is processed by visual as well as somatosensory brain areas, which makes it a prime candidate for elucidating the interplay between the two systems with respect to the FLE. Secondly, discussions about the mechanisms of the FLE tend to recur to evolutionary arguments, while most studies employ highly artificial stimuli with constant velocities.

Methodology/principal finding: Since biological motion is ecologically valid it follows complex patterns with changing velocity. We therefore compared biological to symbolic motion with the same acceleration profile. Our results with 16 observers revealed a qualitatively different pattern for biological compared to symbolic motion and this pattern was predicted by the characteristics of motor resonance: The amount of anticipatory processing of perceived actions based on the induced perspective and agency modulated the FLE.

Conclusions/significance: Our study provides first evidence for an FLE with non-linear motion in general and with biological motion in particular. Our results suggest that predictive coding within the sensorimotor system alone cannot explain the FLE. Our findings are compatible with visual prediction (Nijhawan, 2008) which assumes that extrapolated motion representations within the visual system generate the FLE. These representations are modulated by sudden visual input (e.g. offset signals) or by input from other systems (e.g. sensorimotor) that can boost or attenuate overshooting representations in accordance with biased neural competition (Desimone & Duncan, 1995).

Figure 1. Stimulus type.

Start positions for the biological (A) and the symbolic (B) clips in the 1^st person perspective (cf. Fig. 3).

Figure 2. Motion velocity profile.

Velocity is measured in degrees per second (deg/s) and is plotted for each of the 44 movie frames. A constant velocity as usually employed in FLE research would be represented by a line parallel to the x-axis. The depicted profile was derived from the recorded ‘real’ biological movement (i.e. a person reaching for a cup in top-view, cf. Fig. 3) and it was identical for the biological and the symbolic motion conditions in the experiment. Note that this profile only represents the task-relevant motion in the vertical direction (cf. Fig. 3). Vertical lines in the graph denote the frames in which the flash could occur – the thicker line is lag 0 (cf. Fig. 3). In the flash-terminated trials the motion was aborted after the respective frame with the flash. Further explanations in the text.

Figure 3. The flash and the lags.

The Figure shows example stimuli for lags of −6, 0, and +6 frames (columns) for the biological (rows 1 and 2) and the symbolic (rows 3 and 4) stimuli and for the 1^st person (rows 1 and 3) and the 3^rd person (rows 2 and 4) perspective. In the 1^st person perspective always the right hand of the actor moved, while in the (mirrored) 3^rd person perspective always the left hand moved in order to optimise resonance with the right hand of the observer during tracking. Further explanations in the text.

Figure 4. Psychometric functions for subject 12.

(A) Biological Motion (B) Symbolic motion. Each graph shows the frequency of “behind” responses (y axis) against the flash lag (ms; x axis). The individual data points are fitted by cumulative Gaussian psychometric functions for each of the 8 biological (A) and 8 symbolic (B) motion conditions. The box and error bar at each 50% point (PSE) indicate 95% and 99% Confidence Interval respectively. The steepness of the curve, expressed as the SD, gives an indication of the discrimination performance (or JND). Solid line: continued motion, dashed line: flash-terminated motion. Magenta: “1st person no tracking” condition; Red: “3rd person no tracking” condition; Cyan: “1st person tracking” condition Blue: “3rd person tracking” condition; Further explanations in the text.

Figure 5. Motion main effect.

Negative values indicate an FLE in the continuous motion condition while positive values indicate a ‘flash-lead effect’ (FleadE) in the flash-terminated condition. Further explanations in the text.

Figure 6. Interaction between stimulus type, perspective, and agency.

The y-axis expresses the PSE in ms (compare Fig. 4). Biological motion is shown in the left and symbolic in the right graph. Vertical bars denote the 95% confidence interval. Further explanations in the text.

Figure 7. Flash-terminated and continuous trials.

The x-axis shows the four combinations of perspective (1^st vs. 3^rd person) and agency (tracking vs. no tracking). The y-axis expresses the PSE in ms. A) The positive values in the flash-terminated trials reveal that we did not observe an FLagE (negative values) but a flash-lead effect (FLeadE, positive values) in the flash-terminated conditions due to the nature of our stimuli (see Methods). However, the condition that generates the strongest motor resonance (biological, 1^st person perspective), yet without the illusion of control (no tracking) shows the smallest FLeadE (i.e., more of an FLE than the other conditions). Two asterisks indicate p<.01. B) Predominantly negative values are observed in the continuous motion conditions. Further explanations in the text.