(A) Net gravito-inertial acceleration (GIA; thick lines) and net GIA error (thin lines) aligned to start and end of trial, for the vestibular and combined conditions across participants (average across trials over all time constants).
The dashed line represents a conservative choice of the vestibular motion detection threshold according to the relevant literature (8 cm/s2; Kingma, 2005; MacNeilage et al., 2010; Zupan and Merfeld, 2008). Gray region represents the target presentation period. Shaded regions denote ±1 SEM. (B) Net tilt velocity aligned to start and end of trial, for the vestibular and combined conditions across participants. Dashed line represents the estimated tilt/translation discrimination threshold of 1 deg/s: although tilt/translation discrimination thresholds have not been explicitly studied, we can use the rotation sensation thresholds of the semicircular canals to estimate what that threshold would be. Since it is the rotation velocity that tells a participant that they are tilting and not translating, we propose that the tilt/translation discrimination threshold is at least the same as the rotation sensation threshold (if not larger; Lim et al., 2017; MacNeilage et al., 2010). Shaded regions represent ±1 SEM across participants. Inset shows the probability distribution of displacements during the suprathreshold tilt period after trial onset (~0.6 s). Although the tilt can be perceived by the participants during trial onset, the displacement during that period does not exceed 10 cm and could potentially not contribute significantly to steering errors, for three reasons: (a) the displacement during that period is negligible, (b) tilt velocity is kept below the perceptual threshold for the remainder of the trajectory, (c) GIA is always above the motion detection threshold of the vestibular system. However, since the initial tilt could be perceived (as it briefly exceeded the canal detection threshold), this might alter the perceived orientation of the participants. In turn, this could influence the extent to which vestibular cues would be used as input to the path integration system (see Discussion ‘Limitations and future directions’ for further discussion). Thus, perceived tilt might be used as an indicator of trial onset, but it cannot contribute to path integration for three reasons: (a) the displacement during that period is negligible, (b) tilt velocity is kept below the perceptual threshold for the remainder of the trajectory, (c) GIA is always above the motion detection threshold of the vestibular system.