Visual agnosia

Much of the research in this area has centred on visual agnosia, and our coverage of the experimental evidence focuses on this disorder. Connectionist models designed to account for some of the major perceptual disorders have recently been put forward, and are discussed in the next major section of the chapter.

Two tests used to assess apperceptive agnosia are the Gollin picture test and the incomplete letters task. In the Gollin picture test, the participants are presented with a series of increasingly complete drawings of an object. Patients with apperceptive agnosia require more complete drawings than normal individuals to identify the objects. The incomplete letters task involves presenting letters in fragmented form and asking the participants to identify them. Patients with apperceptive agnosia are worse than normals at this task. Patients with apperceptive agnosia perform worse than those with associative agnosia on tests involving matching and copying objects that patients cannot name (see Kohler & Moscovitch, 1997).

Warrington and Taylor (1978) argued that the key problem in apperceptive agnosia is an inability to achieve object constancy, which involves being able to identify objects regardless of viewing conditions. They tested this hypothesis using pairs of photographs, one of which was a conventional or usual view and the other of which was an unusual view. For example, the usual view of a flat-iron was photographed from above, whereas the unusual view showed only the base of the iron and part of the handle. When the photographs were shown one at a time, the patients were reasonably good at identifying the objects when they were shown in the usual or conventional view, but were very poor at identifying the same objects shown from an unusual angle.

Warrington and Taylor (1978) obtained more dramatic evidence of the perceptual problems of these patients when they presented pairs of photographs together, and asked the patients to decide whether the same object was depicted in both photographs. The patients performed poorly on this task, indicating that they found it hard to identify an object shown from an unusual angle even when they knew what it might be on the basis of their identification of the accompanying usual view.

The findings obtained by Warrington and Taylor can be explained by assuming that the patients found it hard to transform unusual views of objects into appropriate 3-D model representations as described by Marr (1982). However, the view of an object can be unusual in at least two ways. It can be unusual because the object is foreshortened, thus making it hard to determine its principal axis of elongation, or because a distinctive feature of the object is hidden from view.

These possibilities were compared by Humphreys and Riddoch (1984, 1985). They used photographs in which some of the unusual views were based on obscuring a distinctive feature, whereas others were based on foreshortening. The participants either had to name the object in a photograph, or they had to decide which two out of three photographs were of the same object.

In four patients having right posterior cerebral lesions, Humphreys and Riddoch (1984, 1985) found that they performed poorly with the foreshortened photographs but not with those lacking a distinctive feature. Marr and Nishihara (1978) argued that foreshortening makes it especially hard to attain a 3-D model representation, and so the findings are generally consistent with their theoretical position.

Patients with associative agnosia have problems in naming objects. However, they are fairly good at copying and matching objects they cannot name. For example, they can match photographs of objects taken from unusual angles. Some associative agnosics can discriminate on the object decision task between perceptually similar objects such as pictures of real objects and artificial objects created by switching the parts of real objects (e.g., Sheridan & Humphreys, 1993).

Some patients with associative agnosia show the phenomenon of category specificity, meaning that they have special problems in recognising certain categories of objects. For example, Warrington and Shallice (1984) studied a patient, JBR, who suffered from severe associative agnosia. He had much greater problems in identifying pictures of living than of non-living things, having success rates of about 6% and 90%, respectively. The findings from other studies indicate that the pattern shown by JBR is much more common than the opposite pattern, i.e., worse recognition of nonliving than of living things. However, Warrington and McCarthy (1994) did report on one patient who showed consistently worse performance with drawings of objects than with drawings of animals. The task involved deciding which of five drawings was most closely associated with the target drawing.

How can we account for these findings? The greater difficulty in recognising living than nonliving things can be explained by assuming that pictures of living things are more similar to each other than are pictures of non-living things, and are thus harder to recognise. Evidence consistent with this view was reported by Gaffan and Heywood (1993). They asked normal individuals to name pictures of living and non-living things that were presented for only 20 ms each. The key finding was that all the participants performed much worse on living than on non-living things, indicating that living things are harder to recognise.

The findings of Gaffan and Heywood (1993) do not explain why a few patients with associative agnosia have greater difficulty in object recognition for non-living objects than living ones. Perhaps different brain areas contain at least some of the semantic knowledge used in recognising living and non-living objects. A theory of this type was put forward by Farah and McClelland (1991), and is discussed in the next section.

An interesting case of agnosia was reported by Humphreys and Riddoch (1987). They studied HJA, who could not recognise most objects after suffering a stroke. However, he produced accurate drawings of objects he could not recognise, and he could draw objects from memory. His perceptual problems seem to centre around the fact that he found it very hard to integrate visual information about the parts of objects in order to see the objects themselves. In HJA's own words: "I have come to cope with recognising many common objects, if they are standing alone. When objects are placed together, though, I have more difficulties. To recognise one sausage on its own is far from picking one out from a dish of cold foods in a salad" (Humphreys & Riddoch, 1987).

Evidence that HJA had a serious problem in grouping or organising visual information was obtained by Humphreys et al. (1992). The task of searching for an inverted T target among a homogeneous set of distractors (Ts) is easy for most people. However, HJA's performance was very slow and error-prone, presumably because he found in very hard to group the distractors together.

HJA is not the only agnosic patient to have problems with integrating visual information. For example, Behrmann, Moscovitch, and Winocur (1994) studied CK, a man who suffered head injury in a car crash. CK was reasonably good at copying a figure consisting of three touching geometric shapes (two diamonds and a circle). Nearly all normal individuals would copy this figure object by object. What CK did was to follow the outer boundary of the whole figure, which meant that he often moved on to the next shape before completing his drawing of the last one. As Gazzaniga, Ivry, and Mangun (1998, p. 193) concluded, "An inability to integrate features into a coherent whole may be the hallmark of many agnosic patients."

In sum, as Humphreys and Riddoch (1993) pointed out, the distinction between apperceptive agnosia and associative agnosia is oversimplified. According to them, visual object recognition involves a series of stages: feature coding; feature integration; accessing stored structural object descriptions; and accessing semantic knowledge about objects. Problems with visual object recognition can occur because of impairments at any of these stages. This is a more complex (but realistic) position than the simple distinction between apperceptive and associative agnosia.

Business Correspondence

Business Correspondence

24 chapters on preparing to write the letter and finding the proper viewpoint how to open the letter, present the proposition convincingly, make an effective close how to acquire a forceful style and inject originality how to adapt selling appeal to different prospects and get orders by letter proved principles and practical schemes illustrated by extracts from 217 actual letter.

Get My Free Ebook


Post a comment