166 |
Georg W. Alpers and Antje B.M. Gerdes |
|
|
2. "Blind" in One Eye: Binocular Rivalry
Information from the environment reaches the two eyes independently. Under normal circumstances this information is combined into a meaningful spatial representation. However, conscious perception does not inevitably represent the physical environment. Instead, perception is the end product of several steps of selective processing. While many stimulus properties are still processed by the eyes’ sensory cells, only a small fraction of them reaches conscious awareness. This is especially obvious when ambivalent information is presented to the two eyes and information cannot be combined to a meaningful impression at later stages of processing in the brain. When competing pictures are presented to the two eyes and a distinct impression cannot be evoked, this results in a fascinating perceptual phenomenon called binocular rivalry. For a given period of time, one of the pictures is perceived dominantly while the other is suppressed and thus removed from conscious awareness. An unpredictable alternation between the two impressions ensues. At times rivalry can also result in percepts which consist of mixtures of both pictures combined of parts from both. During extended periods of time, input from one eye is completely suppressed from conscious awareness. Thus, perceptual changes occur while visual input remains constant.
Binocular rivalry is a remarkable phenomenon and offers the possibility to further investigate visual perception. Importantly, conscious control over what is perceived during binocular rivalry is very limited (Meng and Tong, 2004; Tong, 2001). In general, binocular rivalry enables the researcher to investigate features of conscious perception, and processes underlying perception, in detail.
2.1. What Helmholtz Knew Already
Recently, binocular rivalry receives growing attention in research focused on consciousness, both from the psychological and the neurological point of view, but it is not a newly discovered phenomenon. Binocular rivalry has been a well known phenomenon for a very long time (see Humphrey and Blake, 2001). In 1593 already, Porta reported a perceptual phenomenon which occurred when he held the two pages of a book right in front of his two eyes (cited by Wade and Ono, 1985). He reported being able to temporarily read one of the pages, while the other was invisible.
Early systematic investigations of binocular rivalry trace back to Wheatstone (1838). For his experimental investigations he developed the mirror stereoscope –
Temporarily Blind in One Eye |
167 |
|
|
an optic apparatus which makes it possible to present different pictures to the two eyes (see a modern version in Figure 3).
Ever since binocular rivalry has been studied scientifically, the underlying neuronal processes have been discussed controversially. In spite of the multitude of interesting findings, the neuronal mechanisms have not yet been unambiguously clarified. The main disagreement that prevails is whether the competition for conscious perception takes place at very early or later stages of visual processing. One of the earliest theories was introduced by Helmholtz (1924) and many other studies are based on it.
He assumed that the eyes’ visual fields are completely independent of each other and that under normal circumstances the consistent perceptual impression does not occur until higher mental processes have taken place. Consequentially, he called the phenomenon “retinal rivalry”. According to this theory, the decision about dominance or suppression would only occur after both pictures have been processed independently and higher mental processes such as attention would select among the two.
Figure 3. The mirror stereoscope used in our laboratory (sample pictures on the computer screen).
168 |
Georg W. Alpers and Antje B.M. Gerdes |
|
|
Hering (1886), on the other hand, favoured another model. He assumed that early inhibitory interactions in visual processing account for the occurrence of rivalry. This has also been labelled the “low-level” theory. Thus, according to this theory, the decision about dominance or suppression takes place before the two pictures are completely processed.
2.2. Competition between Input from the Eyes or between the Percepts?
This 19th century controversy continues to the present day. According to the “low-level” theory (advocated for example by Blake, 1989), binocular rivalry exhibits typical properties of early visual processes. Therefore, this theory is also called “eye-rivalry” theory. Rivalry thus occurs by means of inhibitory connections between the monocular processing channels. Evidence for this theory comes from imaging studies showing that rivalry alters activity early in the visual stream, e.g., in V1 (Polonsky, Blake, Braun, and Heeger, 2000; Tong and Engel, 2001) and the lateral geniculate nucleus (Haynes, Deichmann, and Rees, 2005). Importantly, processing in these circuits is generally thought to be mostly independent from input from the two eyes. Binocular rivalry suppression of one channel would thus mean that input from one eye is not thoroughly processed before being suppressed.
In contrast, the “high-level” theory postulates that binocular rivalry arises because of a competition between stimulus information, independent from the source of this input. Thus, this perspective is also called the “stimulus-rivalry” perspective. It assumes that rivalry is decided after primary processing in monocular channels is integrated in binocular channels, that is, the crucial processes are thought to be independent from the fact that this input originated from one eye or the other (Logothetis, Leopold, and Sheinberg, 1996). Thus, according to this theory, rivalry takes place between integrated stimulus representations, rather than information from the two eyes.
Support for this perspective comes from single cell recordings in animals which show little evidence for rivalry-correlated activation in V1, inconsistent effects for visual areas V4 and MT, but clear effects in the inferior temporal cortex. These neurophysiological findings favour the theory that rivalry is the result of a competition between incompatible stimulus representations in higher areas of visual processing, i.e. after information from the two eyes has been integrated in V1 (Sheinberg and Logothetis, 1997).
Temporarily Blind in One Eye |
169 |
|
|
Some findings from human studies also account for the fact that a certain amount of processing has taken place before the competition is decided. For example, suppressed stimuli can produce after-images(e.g., O'Shea and Crassini, 1981). However, the strongest argument for the “high-level” theory is the finding that fragments of a picture presented to one eye each can be reassembled to a consistent percept in conscious perception (Kovacs, Papathomas, Yang, and Feher, 1996). Rivalry then occurs between combined percepts and not between input from one eye. Interestingly, such rivalry between percepts even occurs when parts of the pictures are intermittently projected to one or to the other eye (Logothetis, et al., 1996).
Because convincing support has been found for both theories, many authors conclude that binocular rivalry may involve stages of visual processing (Nguyen, Freeman, and Alais, 2003; Wilson, 2003). From this point of view, the theories are not mutually exclusive, but rather characterize two extremes on a continuum. Rivalry considerably suppresses activation in the primary processing channels, but enough information of the suppressed picture can proceed to brain circuits which process integrated information from both monocular channels.
2.3. Possible Influences from Non-visual Neuronal Circuits
Taking into account the projections from the amygdala to primary sensory areas of the visual cortex (V1, V2) which we mentioned above (Amaral, et al., 2003), it becomes apparent that this may be an avenue for a picture's emotional significance to influence processing within visual circuitry. If the relevance of a stimulus has been detected within subcortical circuits (“low road”) this may lead to more intense visual processing at several later stages of visual perception.
Indeed, it has been shown that the amygdala is more strongly activated by fearful faces, even when they are suppressed in binocular rivalry. Thus, although the emotional material was not available to conscious perception (confirmed by an absence of activation in specialized face-sensitive regions of the ventral temporal cortex), emotional circuitry was (Pasley, Mayes, and Schiltz, 2004; Williams, Morris, McGlone, Abbott, and Mattingley, 2004).
Thus, we proposed that such preferential processing of emotional pictures may result in their predominanceover neutral pictures in binocular rivalry. Because conscious control over binocular rivalry is probably not possible (Meng and Tong, 2004) we argued that demonstrating such predominance would provide particularly convincing evidence for preferential processing during prolonged perception.