Segall, Campbell and Herskovits (1963) measured the magnitude of the above illusions in urban and tribal participants in USA, Africa and the Philippines. They found that urban participants were more prone to the illusions then tribal participants. Results for the Vertical-Horizontal illusions were mixed. This supports the view of the constructionists that perception is based on an internal model of sense-data and is subjective to each person, this is demonstrated by the variance between cultures.
However, Coren (1989) and Pollack (1989) criticises illusion studies between cultures as lacking reliability. Coren and Pollack claim that different racial groups vary in terms of pigmentation of the lens, iris and fundus. They claim that these optical factors produce variations in illusion magnitude and picture perception.
If direct realists were to maintain their argument they would need to ignore developments in the senses as age passes, for example, Atkinson (2000) and Owsley, Sekuler and Siemsen (1983) found that the ability to resolve fine spatial detail (visual acuity) in an infant does not reach adult levels until 4 or 5, the levels of increased ability then evens out but begins to gradually worsen at approximately 30 years of age. However, a 90 year olds visual acuity is still five times that of a 6 month old. This is incompatible with the claim that the world is as we see it (that of the naive realists) as if this were true, as one ages the world, and basic objects would change significantly.
However, the naive realists would claim that this was due to physical developments in a child; naive realists would claim that an infant obviously cannot perceive as well as a full grown adult as their eyes are less mature. Atkinson (2000) found that the packing of foveal cones in the retina does not reach adult levels until 4 years, furthermore, Youdelis & Hendrickson (1986) found that the length of the foveal cones outer segment increases by a factor of 10 during infancy. Foveal cones are essential for the retinas sensitivity to light can begin to explain an infants inferior vision. Infants also have smaller eyes, this would also limit their visual capabilities.
However, Banks and Bennett (1988) claim that only half of the difference between infants and adults acuity can be attributed to factors in the retina, the rest reflect limitations in post-receptoral neural processing. This becomes strong support for the indirect realists claim that post sensory processing plays a great influence in perception, in fact, Banks and Bennett demonstrate that post sensory processing is as important as physiological development in an infant.
If, as direct realism states, perception is a bottom-up process then on a simple perception task, 100%, or almost 100% of participants would be able to recognise a simple three dimensional image, Hudson (1960) tested this on black and white, schooled and unschooled children in South Africa by showing participants a drawing of a hunter aiming at an antelope with an elephant in the background and asking them to identify what was shown in the picture. Hudson found that in black participants 8.56% of unschooled and 66.2% of schooled participants correctly identified the picture as 3-D whilst in white participants the figures were 15.33% and 58.61% respectively. This lead to Hudson concluding, “formal schooling and informal training combined to supply an exposure threshold necessary for the development of the process [3-D perception]. Cultural isolation was effective in preventing or retarding the process.” From this, it follows that perception in vision is a top-down process and supports the constructivist view; it is shaped by schooling and cultural interaction and without either three-dimensional perception will diminish or not develop.
However, Hagen and Jones (1978) reviewed this study and claimed that Hudson’s pictures were impoverished and geometrically inaccurate line drawings, Gagen and Jones claimed that the drawings lacked textural detail s and grey level gradation. They highlighted that even one in three responses from schooled subjects were incorrect, this adds a serious question to whether the results were due to unreliable methodology.
In conclusion, there are strong implications for the naive realists; mainly that it is very difficult to ignore a mediating agent between sensory perception and analysis by the brain preventing an interpretation based fully on the external world. However, it is also true that the constructionists also represent an extreme; if the constructionists views were accepted then we would live in a Matrix style world where we have no experience of the outside world. There is also the question of how the brain and external world interact so well if the senses are so unreliable. A more likely solution is that there is a medium between these two extremes, perception is not fully one or the other, but between both theoretical perspectives to allow a greater interaction between the senses and the brain, but without taking away from the idea of sense-data.
References
Gordon, I. E. (1989/1997/2004). Theories of visual perception, third edition, chapter 7, Chichester: Wiley.
Le Morvan, P., Arguments against direct realism and how to counter them. Retrieved: 13/12/2008, from: www.tcnj.edu/~lemorvan/DR_web.pdf
Mather, G. (2006). Foundations of perception, chapter 13, Hove: Psychology Press.
Preston, J Naïve Realism and the Argument from Illusion. Retrieved: 13/12/2008, from:http://www.rdg.ac.uk/AcaDepts/ld/Philos/jmp/Theory%20of%20Knowledge/NaiveRealism.htm