The parafovea may refer to a position at a visual angle of approximately 4° from the normal fixation point or to the region extending to that angle from the fovea. It is compositionally distinguished from the fovea by its lower cone and ganglion cell densities and by its higher rod density. As may be expected, these differences affect its visual perception characteristics. Its spatial contrast sensitivity, visual acuity, and other visual attributes are lower than those of the fovea, but, as shown by its critical fusion frequency (CFF), its temporal frequency resolution extends to higher frequencies than that of the fovea. Its color perception is essentially the same as that of the fovea, but under low luminance, it becomes similar to tritanopic diachronic perception. These characteristics can be inferred from the relative distribution of the three cone types in the parafoveal retina. In reading, moreover, parafoveal information contributes to line of vision movement and to fine foveal information processing.
In the parafovea, the cone density declines from the high concentration of cones in the fovea, the rod density rises from the complete absence of rods in the fovea , and the density of ganglion cells, which form the retinal outlet of the optic nerve, declines together with the ganglion cell layer thickness from the high density and thickness of the foveal region [4, 5]. Although the parafovea and the fovea both form part of the macula, they differ significantly in macular pigment density [6, 7]. These structural and compositional differences suggest that they affect parafoveal visual perception.
Spatial contrast sensitivity, a fundamental characteristic of visual perception, is lower in the parafovea than in the fovea. The parafoveal contrast at an eccentricity of 4° is functionally characterized as a spatial frequency band-pass filter, like that of the fovea at 0°, but with lower overall sensitivity [8, 9]. In the region most sensitive to spatial frequencies of 2–8 cycles/°, the parafovea sensitivity is approximately 1/2–1/3 of the foveal sensitivity . Visual acuity is also significantly lower than in the fovea. In Landolt ring visual acuity measurements, subjects with a foveal visual acuity of 2.0 exhibit a sharply lower visual acuity of 0.3 in the parafoveal region at approximately 4° , thus showing that a larger character size is necessary for character deciphering. It has been found that the character size threshold for deciphering increases linearly throughout the foveal region and on out to an eccentricity of 30° . In temporal frequency contrast sensitivity, on the other hand, the parafoveal resolution extends to a higher frequency than the fovea, as measured by CFF (critical fusion frequency) .
Parafoveal color perception does not depart from that of the fovea under relatively high illuminance, but under low illuminance, it resembles tritanopic dichromatic perception (small-field foveal vision) . This may be inferred from psychophysical measurements that show nearly the same L- and M-cone ratios in the fovea and the parafovea  but extremely low S-cone distribution in the parafovea as compared with its L- and M-cone distributions .
Three phenomena that have been observed in reading experiments concerning parafoveal information processing are known as skipping, parafoveal-on-foveal (PoF), and preview-benefit effects. The skipping effect consists of skipping every third word or so, which would presumably not be possible without identification of the word via the parafovea. The skipped words are generally short terms such as “to,” “for,” “in,” and “the,” which supports the view that a word’s lexical status contributes to the determination of whether it will be skipped.
In the PoF effect, the occurrence of a parafoveal word that is difficult to understand affects the foveal fixation and thus the field of vision. As noted above, parafoveal processing affects foveal processing, and a low-level PoF effect has been well established in experiments involving the occurrence of unfamiliar character combinations in the parafoveal region. The preview-benefit effect has been observed in experiments in which the word on the parafovea that is subsequently to be subjected to foveal processing is replaced by a related word during the saccade. If the replacement word is related to the original parafoveal word, the fixation on that word by the fovea is shorter and its processing thus faster than if the replacement word is unrelated. Recent studies have indicated that the preview-benefit effect occurs in sentences independently of the initial sentence word, mid-sentence words, and final sentence word.
These three effects are currently the subject of intense discussion and further findings may be expected.
- 1.Polyak, S.L.: The Retina. University of Chicago Press, Chicago (1941)Google Scholar
- 2.Osaka, N.: Psychophysical Analysis of Peripheral Vision. Kazama Shobo Press, Tokyo (1983). (in Japanese)Google Scholar
- 3.Wandell, B.A.: Foundations of Vision. Sinauer Associates, Sunderland MA, US (1995)Google Scholar
- 5.Iwasaki, M., Inomata, H.: Relationship between superficial capillaries and foveal structures in the human retina. Invest. Ophthalmol. Vis. Sci. 27, 1698–1705 (1986)Google Scholar
- 6.Gegenfurtner, K.R., Sharpe, L.T. (eds.): Color Vision from Genes to Perception. Cambridge University Press, Cambridge, UK (1999)Google Scholar