Handbook of Cognitive Neuropsychology

What Deficits Reveal About the Human Mind

Brenda Rapp

• Publisher: Psychology Press
• Year: 2015
• ISBN:
  9781317710233
• Edition: 1
• Topic: Psychologie
• Subtopic: Histoire et théorie en psychologie

This volume reviews the full range of cognitive domains that have benefited from the study of deficits. Chapters covered include language, memory, object recognition, action, attention, consciousness and temporal cognition.

Part 1
Foundations

The most beautiful thing we can experience is the mysterious. It is the source of all true art and science.

—Albert Einstein

1
Assumptions and Methods in Cognitive Neuropsychology

Max Coltheart

WHAT IS COGNITIVE NEURO PSYCHOLOGY?

Cognitive neuropsychology is a branch of cognitive psychology.

Cognitive psychology is that science which seeks to learn more about the nature of the mental processes responsible for our ability to perform such basic cognitive activities as understanding and producing language, recognizing objects and people, storing information in memory and subsequently being able to retrieve it, acting intelligently upon the physical world, and so on. Cognitive psychologists also interest themselves in higher-level cognitive processes such as reasoning and problem-solving, the formation of beliefs about the world and about other people and how we assess the plausibility of such beliefs once they have occurred to us, and the cognitive nature of social interaction.

The most common way of investigating cognition is to study people with normal cognitive abilities as they are performing some particular cognitive task; the cognitive psychologist develops a theory about the nature of the mental processes used for performing this task, makes deductions from this theory about what results should be observed in some experiment involving that kind of cognition, carries out such an experiment, and considers whether the data confirm or falsify the theory.

Another way of doing cognitive psychology is to collect data relevant to the theory from people who have disorders of cognition. Any such disorder might be acquired or it might be developmental. If someone in whom a particular cognitive function had been normal suffers damage to the brain which specifically impairs that form of cognition that person has an acquired disorder of cognition. In contrast, someone who has not been able to acquire the cognitive ability to a normal degree has a developmental disorder of cognition. These disorders of cognition are relevant to theories about normal cognition, because such theories make predictions not only about the results of experiments with cognitively intact people, but also about ways in which brain damage can impair cognition, and about ways in which the acquisition of a cognitive ability can go awry. Thus, data from people with acquired or developmental disorders of cognition can confirm or falsify theories about cognition; this way of doing cognitive psychology is cognitive neuropsychology.

For example, suppose one believed that our ability to recognize all kinds of visual stimuli—objects, faces, and the printed word—used a single common visual recognition system. This theory predicts that if in a particular person brain damage has impaired the ability to recognize visually-presented objects, such a person will also have an impairment in the ability to recognize faces and an impairment in the ability to recognize visually-presented words. If experimental investigations of this person reveal that face recognition and visual word recognition are both intact, that falsifies the original theory about visual recognition. The same would be true if one found, in a person who had no history of brain damage, that as he grew up he had acquired the ability to recognize faces and printed words but was never able to acquire the ability to recognize visually-presented objects.

Cognitive neuropsychology is not a kind of neuropsychology (even though it is highly relevant to neuropsychology) because, to put the matter in a nutshell, cognitive neuropsychology is about the mind, while neuropsychology is about the brain. Consider the hypothetical patient described above, who after brain damage could no longer recognize objects but could still recognize faces and visually-presented words. Presumably that means that there must be a region of the brain that is needed if we are to recognize objects, but not needed for recognizing faces or printed words. One might use brain-imaging methods with this patient to try to discover where in the brain this objectrrecognition system is located. However, whether one succeeded or failed in the attempt to locate such a region in the patient’s brain is simply irrelevant to the conclusion drawn about the cognitive organization of visual recognition on the basis of the patient’s performance. The theory that objects, faces, and visually-presented words are all recognized by a single common visual recognition system is still falsified by the finding of a person in whom brain damage affected only object recognition, regardless of what can or can’t be discovered from studying the damaged brain of such a person.

Cognitive neuropsychology is also not necessarily concerned with treatment or rehabilitation, even though it might be relevant, even highly relevant, to them. Suppose one became convinced that all of the kinds of acquired or developmental cognitive disorders studied by cognitive neuropsychologists were completely untreatable—convinced, that is, that a child with a particular developmental disorder will never be able to acquire that cognitive ability, no matter what treatment is used, and that a brain-damaged patient who has lost a particular cognitive ability will never be able to regain it, no matter what treatment is used. Even if all of this were true, that would have no implications at all for the practice of cognitive neuropsychology. Data from people with cognitive disorders could still be used to confirm or falsify theories about normal cognition. On the other hand, it might be true one day—though it isn’t true at present-that our theories about how normal cognitive processes operate have become so detailed that they make predictions about whether particular treatment techniques will or won’t be effective. If that day comes, then treatment studies will be relevant to theories about normal cognition, and cognitive neuropsychology will be directly concerned with issues involving treatment.

Of course, there is already a way in which treatment studies can yield cognitive-neuropsychological evidence, since sometimes theories make predictions about the outcomes of treatment studies. If whether or not function Y depends upon function X is a question of relevance to a theory, and if a patient has an impairment of function X, and if successful treatment of function X is achieved, then one should observe an improvement in function Y. Failure to observe this would count as evidence against the theory.

Finally, although cognitive neuropsychology is distinct from clinical neuropsychology, it is important for clinical neuropsychology in a number of ways. For example, highly sophisticated methods of cognitive assessment can be developed if one uses as a starting point a detailed theory of the relevant cognitive system. The PALPA (Psycholinguistic Assessments of Language Processing in Aphasia) battery for assessing disorders of language (Kay, Lesser, & Coltheart, 1992) and the BORB (Birmingham Object Recognition Battery) for assessing disorders of visual perception and visual recognition (Riddoch & Humphreys, 1993) are good examples of such theory-based assessment methods. Cognitive-neuropsychological assessment methods such as these allow the clinician to pinpoint precisely which aspects of cognition are still intact and which are impaired in particular patients, and this provides a specific focus for treatment and rehabilitation.

THE CASE OF AC

Investigations of the patient AC by Coltheart et al. (1998) provide a useful example that provides a more detailed illustration of just how cognitive neuropsychologists seek to learn more about the nature of normal cognition from investigations of people with impairments of cognition.

AC was a 67-year-old man at the time of these investigations, and had formerly been employed as a clerical worker with the public railway system in New South Wales, Australia. He had a history of cardiovascular and cerebrovascular disease, and a CT scan performed four days after he had suffered a stroke revealed a recent lesion in the territory of the left middle cerebral artery, plus a number of older small lesions in both cerebral hemispheres indicative of earlier more minor cerebrovascular incidents.

Assessment of AC’s linguistic and other cognitive abilities revealed some impairments that were surprisingly severe given the relatively minor damage evident in the CT scan. For example, AC’s reading ability was almost completely abolished; he could not even judge that A and a are two forms of the same letter while A and e are not, a very elementary reading task. This was not because of some impairment of vision, since he could judge that A and A are the same while A and E are not, and could copy such letters correctly, indicating that he could see perfectly well. And since he had been a clerk, it is not likely that his severely impaired ability to read would have been present prior to his stroke. AC’s writing was also almost completely abolished; he could write almost nothing to dictation except his name and address, and when shown an uppercase letter and asked to write its lowercase form next to it, he could not do so. This was not a motor problem, because he was entirely capable of copying an uppercase letter in its uppercase form, as mentioned above.

AC’s ability to generate spoken words, in conversation or when trying to name a picture, was also severely affected; he knew what he wanted to say, but could rarely find the actual word he wanted to produce. He was also very poor at copying pictures and at drawing to dictation.

Studying his reading, his writing, or his drawing could well have provided interesting new information about how these cognitive activities are normally carried out. However, a conversation with AC involving his knowledge about the properties of objects—in particular, animals—led to a series of studies of his knowledge of word meanings. This conversation was prompted by a paper with the title “The Oyster with Four Legs” (Sartori & Job, 1988) which had appeared just before we first met AC. The conversation went as follows:

MC: How many legs does an oyster have?

AC: A few.

MC: I see. What about an ant?

AC: Some.

MC: A caterpillar?

AC: No legs.

MC: What about a snake?

AC: None.

MC: And a seagull?

AC: Four legs.

This inability to provide the required information about animals was formally tested by choosing a set of 18 animals, 9 possessing legs and 9 not, and asking AC to respond “Yes” (i.e., it has legs) or “No” (it hasn’t) when the names of these animals were spoken to him. Chance performance would be 9 out of 18; AC scored 10 out of 18.

Why did he have such difficulty performing this task? One possibility is a category-specific loss of semantic information where the affected category is animals: This was the case with the patient described by Sartori and Job (1988). That was easy to test: A set of 20 inanimate objects, 10 with legs (e.g., chair, table, sofa, etc.) and 10 without legs, was chosen. When asked to indicate whether or not these objects had legs, in response to the spoken names of the objects, AC again failed to perform better than chance, scoring 11 out of 20 correct. So the difficulty here was not with loss of semantic information specifically concerning animals.

Perhaps it had something to do with the word or the concept “leg”? No, because AC was also at chance when the task was to respond to the question “Does it have a tail?”, scoring 12 out of 22 correct and on a subsequent occasion, 11 out of 22 correct with the same items.

Was it just knowledge of the parts of objects that AC had lost? No, because he was unable to make judgements based just on the overall shapes of objects (“Is it round or not?” yielded 15 out of 28 correct) or just on their color (“Is it colored or just black and white?” yielded 12 out of 20 correct).

Well, perhaps what he had lost was information about the perceptual properties of objects; so he was then tested for his knowledge of nonperceptual information about objects. Asked to classify animals as Australian or not, he did very well (18 out of 20 correct; so he knew that a kangaroo is Australian and an elephant is not, despite his inability to report how many legs each has). Asked to classify animals as dangerous or not, he also did very well (19 out of 20 correct), and he also succeeded with the question “Do people usually eat this animal?” (23 out of 24 correct) and “Does this animal live in water or not?” (18 out of 20 correct). These four tests all require access to nonperceptual knowledge about objects, and AC performed all four very well.

His failures on the legs, tails, shape, and color tests could be because he had lost all perceptual knowledge about objects; or was it just specifically visual knowledge which had gone? That was tested by asking for perceptual but nonvisual knowledge. On a test of auditory perceptual knowledge (i.e., “Does it make a sound?”), he scored 24 out of 26 correct; and on a test of olfactory perceptual knowledge (i.e., “Does it have a smell?”), he scored 19 out of 20 correct.

These results support the following conclusion: What happened to AC is that he lost all information about the visual properties of objects, while still retaining knowledge about their nonperceptual properties, and also knowledge about perceptual properties which are not visual.

What does this tell us about how knowledge of objects is represented mentally? If everything we know about objects is represented in a single object knowledge system, it is hard to see how one particular form of knowledge (e.g., visual knowledge) could be lost while other forms of knowledge are still essentially intact. We are therefore led to the view that there is a system of knowledge about what objects look like which is quite separate from other stores of knowledge about other kinds of properties of objects.

Now, it is possible that there are just two systems of object knowledge, one which tells us what objects look like and another which tells us everything else about them (i.e., whether an object has a smell, lives in Australia, makes a noise, lives in water, etc.). But it doesn’t seem very plausible to argue that all the forms of perceptual information are stored together with nonperceptual information, except for visual information, which has its own proprietary store. What seems more plausible, given that we want to say that there’s a system just for visual knowledge of objects, is that there is also a system just for auditory knowledge about objects, and another one just for olfactory knowledge about objects—indeed, a separate system of perceptual object knowledge for each of our senses, plus a system of nonperceptual object knowledge. If so, we should find patients who know that a radio has no legs and a table does, but can’t tell you which of them makes a noise (these hypothetical patients have lost their system of auditory knowledge about objects but still have their system of visual knowledge), patients who know that gasoline isn’t normally drunk and vodka is, but can’t tell you which of these has a smell (loss of olfactory object knowledge), and so on.

A key point here is that, although we have been led to this theory about how object knowledge is stored by studies of someone with impaired object knowledge, the theory is a theory about how such knowledge is represented in the intact person—it is a theory about all of us, including about AC prior to his brain damage. That, of course, is the essence of cognitive neuropsychology: building a theory about normal cognition from a study of abnormal cognition.

Like any other methodology, cognitive neuropsychology makes certain assumptions—that is, there are certain things which need to be true, or at least approximately true, if studies of people with impaired cognition are to be informative about the nature of intact cognitive systems. Those assumptions are discussed below, but before they can be discussed, it is necessary first to say something about a concept that is central for those assumptions—the concept of modularity.

MODULARITY

Let’s suppose the cognitive task we are interested in is the very general task of understanding visual stimuli, whether they ...