Illusions as neuro-signs

Abstract

Almost the whole of visual science is concerned with illusions — if they are defined as images in which the physical description differs from the image perceived. One of the attractions of illusions, apart from their obvious appeal as geometrical abstract art, is that they direct us to an understanding of visual processing at many levels.But before the question of why an illusion is perceived can be addressed, we must have a physical description of the image, which is not always straightforward. After all, what is the physical description of a stick that is partially immersed in water? It is straight if the stick itself is measured, but not if a photograph is taken of it.For this reason (among others) psychologists make a distinction between the distal (or physical) stimulus and the proximal stimulus (its projection onto the retina). An illusion occurs when there is a mismatch between the proximal stimulus and perception.The distinction between distal and proximal stimuli can be appreciated with the image of black and white concentric circles below. The distal stimulus is circular and stable, but its projection onto the retina is distorted and dynamic. Spokes appear to radiate from the centre and to rotate; they are probably a consequence of slight astigmatic changes in the curvature of the lens, because they are reduced when viewed through a pinhole, they have similar temporal dynamics to measured fluctuations in the curvature of the lens, and they do not occur in those who have no crystalline lens (aphakia). Aphakic individuals lacking a lens in one eye see such patterns as stable using that eye, but see it with radiating spokes using the eye that has a lens.Figure 1View Large Image | View Hi-Res Image | Download PowerPoint SlideIn addition, viewing the pattern from a distance will reveal a portrait of Jan Evangelista Purkinje, who first described the radiating spokes phenomenon in 1825. The portrait is carried by low spatial frequency variations in the pattern, which correspond to the coarse grain over the whole pattern; these are normally inhibited by the higher spatial frequencies (the fine grain of the thin lines) when the pattern is close to the eye. When viewed from a distance, the high spatial frequency content (thin lines) is not sharply resolved and the low spatial frequency pattern emerges.Figure 1View Large Image | View Hi-Res Image | Download PowerPoint SlidePurkinje also demonstrated similar effects with radiating lines (see coloured image above). High-contrast coloured or black and white radiating lines are also visually dynamic — they appear to move.The motion consists of shimmering points which are perpendicular to the lines of the distal stimulus. Thus, with a radiating pattern they are circular, and can be seen more readily within the orange annuli. If a white surface is observed following fixation on the pattern for around 30 seconds then scintillating dots will seem to be streaming in a direction orthogonal to the previously presented lines.Regular geometrical patterns produce a range of visual effects; some are based on the optics of the eye (like the spokes) and others (such as the scintillations) could be caused by visual processing in the brain. For example, Purkinje's radiating lines illusion has been examined with the aid of neural imaging. Positron emission tomography (PET) scans indicate that a pattern of radiating lines causes the blood flow in V5 — the area of the brain that is involved in motion processing — to increase. Alternative interpretations, however, relate to eye movement instabilities during fixation. It is likely that several phenomena are induced by these patterns, with a different explanation for each one.When we know that an image is distorted before it reaches the retina — that the proximal stimulus differs from the distal stimulus — we consider that we have some interpretation of the illusion. Thus, the bent-stick distortion is due to differential refraction through air and water. But in general, explanations for the illusion are sought in higher levels of visual processing. Such illusions might provide an index of neural function, which I have called a ‘neuro-sign’.