Concluding remarks

In our daily lives, the omnipresent force of gravity helps the perceptual system to integrate egocentric information arising in the eyes, head, and body into an exocentric frame of reference. With the otolith organs in our inner ear, we possess a sensory system that is specialized to detect gravity and thus to register the orientation of our head in space. Still, the otolith organs do not have the exclusive rights to graviceptive information. The visual and somatosensory system also provide important "down" cues. In normal situations, these sensory systems cooperate as silent partners, making it difficult to determine to what extent each of them is responsible for our sense of orientation. In this thesis I examined the contribution of the otolith system and the visual system to oculomotor and perceptual behavior in two situations where these systems provide discordant information.

First, in the centrifuge experiment described in Chapter 2, the approach was to specifically modify the contribution of the otolith system by exposing subjects to prolonged hypergravity. It was assumed that the vestibular system - the otolith organs in particular - would adapt to this higher G-environment. Accordingly, the otolith system would be "maladapted" to normal gravity for some duration afterwards, thus providing incorrect information about the direction and magnitude of gravity. This view was based on previous observations that centrifugation causes postural imbalance and perceptual changes which may result in motion sickness, similar to the symptoms of the space adaptation syndrome seen during the first days of spaceflight. In the centrifuge experiment presented here, the emphasis was on changes in the otolith-induced ocular torsion response in the need of a quantification of vestibular adaptation. Eye movements were recorded on video tape, and before I was able to "digest" the increasing pile of ocular torsion data in a reliable way, a new automatic method for video-oculography had to be developed first (Chapter 1). Typically, the amplitude of ocular torsion is rather small in man. Hence, using this response in an attempt to relate the perceptual consequences of centrifugation to vestibular adaptation felt like looking for a needle in a hay stack, while in search of the hay stack itself. Nevertheless, the measurements before and after the centrifuge run yielded consistent evidence for otolith adaptation. Additional VOR measurements indicated that this adaptation affected the canal-otolith interaction. Unfortunately I did not observe a direct relationship between oculomotor and perceptual variables. This presumably stems from the fact that the VOR was recorded during passive vestibular testing, while the perceptual consequences were most outspoken during self-generated movements.

The second situation of discordant graviceptive information was created by rotating the visual surroundings of a stationary observer about an earth-horizontal axis (Chapter 4). Static tilt of the visual stimulus resulted in a limited degree of illusory self-tilt, even though the subjects were actually immersed in a tilted environment which was detailed with a rich variety of familiar and polarizing objects. In the case of continuous rotation of the same visual stimulus, however, subjects reported compelling sensations of head-over-heels rotation, irrespective of their actual body orientation. Thus, to completely the override the restraining otolith inputs one needs a rotating polarized scene, i.e. a combination of visual motion and visual polarity information. In the rotating non-polarized sphere of Chapter 5, it was shown that sensations of self-tilt are temporarily enhanced by real body tilt. These effects do not persist after washout in which the body is returned to vertical.

The former two situations were basically concerned with intersensory processes. Chapter 3, on the other hand, dealt with intrasensory interaction within the vestibular system itself. It focused on the relative contribution of the semicircular canals and the otolith organs to the torsional VOR during sinusoidal body roll. Here I learned that there is something peculiar about the torsional component of eye movements. Several workers in the field of eye movements have spent the last century or so to prove that during voluntary saccades the eyes only assume positions which involve no torsion. This restriction is generally known as Listing’s law. All situations which elicit ocular torsion have consequently been labeled as violations of this law. That is bad news for someone who spent more than a year developing a technique to measure ocular torsion! But the case may not be lost yet. It has been argued that the low gain of the torsional VOR in man reflects a conflicting situation for the oculomotor system. On one hand the vestibular system demands to generate a collinear VOR (even if this involves torsion), while on the other hand some "Listing’s operator" demands to avoid torsion. The validity of Listing’s law has essentially been tested during static conditions. The results of Chapter 3 indicate that we should consider the possibility that, during head movements, Listing’s coordinates are dynamically modulated by the otolith system. The otolith-induced component to the response consisted of a modulation of torsional eye position, compensating for head position, while the canal-induced component was clearly charged with the compensation for head velocity. This finding reflects an important functional difference between both vestibular subsystems: the semicircular canals operate in head-centric coordinates, whereas the otolith organs provide a link to exocentric coordinates by registering the orientation of the head relative to gravity. In conclusion, the semicircular canals can be considered as velocity detectors, whereas the otolith organs can be described to be primarily position detectors.

To me, the most remarkable finding is that both the oculomotor and the perceptual responses showed this position-velocity dichotomy. For instance, in the torsional VOR this was visible as the modulation of the beating field (position), superimposed on the slow component (velocity). Similarly, in the rotating sphere, subjects simultaneously experienced a limited degree of self-tilt (position) and a sensation of continuous self-motion (velocity). I believe that the effects observed after centrifugation, or during spaceflight for that matter, can be attributed to the otolith organs providing different position signals than would be anticipated based on previous experience in the normal 1G environment. In this sense, the situations of hypergravity and hypogravity are not essentially different. Until adaptation to the new situation has taken place, which may last from several hours to several days, both sensorimotor and perceptual responses will be inappropriate. It is, however, important to note that adaptation in hypergravity may be accomplished in a different way than adaptation in weightlessness. In hypergravity, there is still a gravitational "down" - which is even stronger than normal - to allow for (intrasensory) adaptation of the otolith system itself. For example, this can be done by reducing the sensitivity, as is suggested by the results of the centrifuge experiment in this thesis. In weightlessness, on the contrary, there is no gravitational "down", and position signals from the otolith organs are completely missing. Here, (intersensory) adaptation may take place by replacing the otolith signals by other sensory signals, for instance from the visual system. No matter how adaptation is achieved, the result will be that the subject is able to function optimally in the new environment. The sensory systems have become silent partners again.