Spatial representation in body coordinates: Evidence from errors in remembering positions of visual and auditory targets after active eye, head, and body movements

Canadian Journal of Experimental Psychology, Mar 2003 by Kopinska, A, Harris, L R

When the head was displaced when viewing the target and then returned (rightwards) to the centre before judging its position, errors were made in the same direction (Figure 13; open circles). This shift is in the same direction as seen in Experiment 2c and is opposite to the shifts in auditory localization found in Experiment 1.

Curiously, when no head movement at all intervened between target presentation, indicating its location, a small but significant shift was also seen (Figure 13; filled circles; -0.020 deg/deg (p

Discussion

This study has shown that the remembered location of lateralized sounds and visual targets are displaced from their correct angular directions relative to the head as a result of the head being in eccentric position with respect to the body but not when the eyes move in the head or the body moves in space.

When the head was actively turned to one side in the dark, the perceived location of remembered auditory lateralized targets shifted in the same direction by approximately 23% (0.023 dB/deg) of the head turn (Figure 6). The percentage was obtained using the calibration procedure described in the methods for converting internally heard sounds to external directions. This shift related to head-on-body position indicates that participants were using head-on-body information in their judgments. Head-on-body information is needed to convert information from a head frame to a body frame, a reference frame conversion which is not required if the location of the sound were stored relative to the head since all judgments were made relative to the head. Furthermore, the head-on-body information that the participants were using was inaccurate: The amount of head displacement was overestimated. When the same displacement of the head on the shoulders was achieved by keeping the head still in space and moving the body (under the participants' control), the error was even larger: 32% (0.032 dB/deg; Figure 7). The amplitude of an active head turn is accompanied by vestibular and neck sensory cues and a sense of effort (efference copy). Moving the body beneath an earth-stationary head by pushing a rotating chair around with the feet is accompanied by sensory cues from the neck but no vestibular cues and the sense of effort is associated with the feet rather than the neck. Greater errors in knowledge of head-on-body position were found when the head's position was monitored by this combination of cues.

The remembered location of visual targets was also displaced during a head movement but by a smaller amount (6%, 0.06 degs/deg; Figures 12 & 13) and in the same direction as the head movement.

FRAMES OF REFERENCE

We postulate an explanation for these findings based on the creation of an internal representation of target locations relative to the body using inaccurately processed head-on-body information. When the head moves under the conditions of these experiments, we hypothesize that the internal representation of the displacement is larger than the actual movement even though the head is perceived to have moved through the correct amount specified in this case by the distance between LEDs. This suggestion is supported by direct measures of perceived head position (Becker & Saglam, 2001), which find an overestimation of head position of between 6 and 18%. Thus if the head is requested to move through 15[degrees], participants can do this and know that they have done it accurately because of feedback - they can see the position of their head-mounted laser. However, the movement feels curiously unnatural. Normal gaze shifts are only partly achieved by a head movement, the short-fall being made up by an eccentric eye position of about 10-15% of the total shift (Becker & Saglam, 2001; Biguer et al., 1984; Gresty, 1974; Kopinska & Harris, 1998; Stahl, 1999; Zambarbieri et al., 1997). When the gaze change is made up entirely by a head movement - leaving the eyes centred in their orbits - we postulate that this results in too large an internal estimate of the head movements' magnitude connected to this unusual situation. When making spatial judgments, the internal representation of a sound is then moved by this too-large amount to a position that is beyond its veridical position. Settings are therefore displaced in the direction of head movement. This is illustrated diagrammatically in Figure 14a. The larger shift found when the head was stationary in space (Experiment 1c, 32%) compared to when it was moving (Experiment 1b, 23%) suggests that vestibular information might contribute to registering the head displacement. If the neck muscles are stimulated alone, illusory motion of a visual target is created (Biguer et al., 1988; Roll et al., 1991; Taylor & McCloskey, 1991), suggesting that vestibular and neck cues combine to create the normal veridical perception of head position in space (see also Mergner et al., 2001). However, when the head and body were moved together in space (Experiment 1d) no perceptual shift was seen (Figure 9).