We describe the characteristics of displacement of the head and hip in norm
al young subjects standing on a moving platform undergoing continuous sinus
oidal horizontal translation in the antero-posterior direction, at frequenc
ies ranging from 0.1-1 Hz. The head, hip and malleolus were marked by light
-emitting diodes (LEDs), and the displacement of each LED was quantified by
(1) the measure of the shift during each cycle of translation, (2) the sta
ndard deviation (SD) of the path travelled during the whole trial, (3) the
power spectrum (PS) of the signal and (4) the cross-correlation (CC) betwee
n pairs of LED signals. At each frequency of translation, with eyes open (E
O), the displacement of head was smaller than that of hip, and the displace
ment of hip was smaller than that of malleolus. With eyes closed (EC), this
order was reversed. The peak value of the CC functions of the pairs malleo
lus/head, malleolus/hip and hip/head decreased by passing from low to high
frequency of translation, under both visual conditions, and decreased more
for the pair malleolus/head than malleolus/hip. The lags between body segme
nt displacements ranged between 30 ms and 150 ms, on average, the former se
gment of each pair preceding the latter. The fast Fourier transformation of
hip and head displacement showed a power spectrum peak at the frequency im
posed by the platform translation. The peak was larger with EC than EO. Wit
h EC, another peak appeared at 0.2 Hz, possibly corresponding to the respir
atory frequency. We conclude that, when vision was allowed, subjects behave
d as a non-rigid, noninverted pendulum, and stabilised head in space. When
vision was denied, head oscillated more than the platform, especially at lo
w translation frequencies. Therefore, the strategy of balance control shift
ed from a pendulum to an inverted-pendulum behaviour, passing from active h
ead-and-trunk control to maximal body compliance to the perturbation.