The various approaches within contemporary cognitive psychology have been discussed at length in the previous chapters of this book. Cognitive psychologists have made much theoretical and empirical headway in making sense of human cognition, especially in recent years. Some of this progress is negative, in the sense that we now know that certain theoretical approaches are actually dead ends. Of course, eliminating erroneous approaches is not the same as discovering the best approach. However, the history of science reveals that it is usually an important step along the way.

Most of the emphasis in this book has been on specific theories and bodies of research. In contrast, the primary aims of this chapter are to provide more global evaluations of the entire approach of cognitive psychology and of its four main perspectives: experimental cognitive psychology; cognitive neuropsychology; cognitive science; and cognitive neuroscience.


The recent dramatic increase in the impact of cognitive neuropsychology, cognitive science, and cognitive neuroscience has led many people to de-emphasise the contribution made by the more traditional experimental cognitive approach. In fact, all three of the newer approaches owe much to experimental cognitive psychology. For example, cognitive neuropsychology became a significant discipline about 20 years after cognitive psychology. It was only when cognitive psychologists had developed reasonable accounts of normal human cognition that the performance of brain-damaged patients could be understood fully. Before that, it was very hard to decide which patterns of cognitive impairment were of theoretical importance. In a similar way, the computational modelling activities of cognitive scientists are often informed to a major extent by pre-computational psychological theories. Finally, the tasks selected by cognitive neuroscientists for their neuroimaging studies are determined by the theoretical and empirical efforts of experimental cognitive psychologists.

A striking success of experimental cognitive psychology has been the way its approach has influenced several areas of psychology. For example, social, developmental, and clinical psychology have all become decidedly more "cognitive" in recent years. Many of the experimental paradigms used by researchers in those areas were initially developed in the research laboratories of experimental cognitive psychologists.

Finally, the methodological contribution of experimental cognitive psychology should not be underemphasised. In cognitive science and the neurosciences, the methodologies for studying phenomena are still being developed. Cognitive modellers are often accused of being unprincipled in their use of models (Cooper & Shallice, 1995), and in brain imaging there are still many issues about the methods used to rule out noise and exclude activation in brain regions that are merely by-products of the focus of a study. In all of this methodological flux, experimental cognitive psychology is the bedrock with its well worked-out empirical methods built up over 100 years of experimentation.


A recurring criticism of experimental cognitive psychology is that it lacks ecological validity, i.e., the findings cannot readily be generalised to the "real" world (see Chapter 8). For example, the participants in most cognitive experiments are well motivated, undistracted, have no other goals competing with that of task completion, and know exactly what they are supposed to do with the task stimuli. The near-optimal conditions in which cognitive psychologists usually assess cognition can be contrasted with those prevailing in the workplace: many office workers have multiple work goals to satisfy, they have to devote time and effort to prioritising goals, they are distracted by phone calls and by people knocking on their door, and their behaviour is greatly affected by social pressures.

The $64,000 question is whether this really matters. That it may do so can be illustrated by an analogy. We could study the limits of people's mobility by having them run races of different distances on a fast running track wearing runningspikes, singlet, and shorts. Measures of some ofthe processes involved could be obtained byrecording heart rate, respiration rate, and so on.However, we would not expect that the information obtained would be of much value when predicting people's speeds while laden down withshopping or walking with their children, nor wouldwe expect to be able to make accurate predictionsof their heart rate or respiration rate under thoseconditions.

In the real world, people are constantly behaving in ways that will have an impact on the environment (e.g., turning on the television to watch a favourite programme). Thus, the responses that people make often change the stimulus situation. In contrast, most of the research of cognitive psychologists involves what Wachtel (1973) referred to as the "implacable [unyielding] experimenter". That is to say, the sequence of stimuli that the experimenter presents to the participant is not influenced by his or her behaviour, but is determined by the experimenter's predetermined plan.

As a result of the implacable experimenter, cognitive psychology is unduly limited in various respects. Consider what cognitive psychologists have discovered about attention (see Chapter 5). Many of the characteristics of divided and focused attention have been identified, but it could be argued that fundamental aspects of attention have been de-emphasised. The focus of attention in most research is determined by the experimenter's instructions. As a result, relatively little is known of the factors that normally influence the focus of attention (but see Chapter 18): relevance of stimuli to current goals; unexpectedness of stimuli; threateningness of stimuli; intensity of stimuli; and so on. This is an important limitation, because it would be impossible to predict someone's cognitive processes and behaviour in most situations without detailed knowledge of the factors determining attentional focus.

Another issue related to ecological validity is what has been called the "decoupling" problem. If a researcher wants to explore some aspect of, say, human memory, then an attempt is usually made to decouple the memory system from other cognitive systems, and to minimise the impact of motivational and emotional factors on performance. Even if it is possible to study the memory system in isolation, it is clear that it usually operates in interaction with other functional systems. Accordingly, the more successful we are in examining part of the cognitive system in isolation, the less our data are likely to tell us about cognition in everyday life. For example, there are influences of emotional states on cognition, and of cognition on emotional states (see Chapter 18). The usual strategy of ignoring emotional factors cannot be recommended if we want to generalise from the relatively unemotional states found in the laboratory to the much stronger emotional states of everyday life.

The key point can be expressed by referring to the distinction between internal validity (the validity of research within the context in which it is carried out) and external validity (the validity of research outside the research situation). Much of cognitive psychology is higher in internal than in external validity. However, this criticism has lost some of its force in recent years. There are several examples in this book of the increased willingness of experimental cognitive psychologists to move closer to "real life". For example, researchers have become more interested in perceptual processing of the human face, which is a very significant stimulus in everyday life (see Chapter 4). In Chapter 5, there is a discussion of attentional and automatic processes in connection with the "real-world" phenomenon of absent-mindedness. Chapter 8 is devoted to everyday memory, Chapter 14 deals with instances of skilled thinking in everyday life, and much of Chapters 11 and 12 is concerned with the important everyday activity of reading.

Some authorities have expressed scepticism about this increased focus on naturalistic research. For example, consider research on everyday memory (see Chapter 8). According to Banaji and Crowder (1989, p. 1190):

We have not been able to see any new principles of memory emerging from the everyday memory studies. Again and again, what seem at first like new, dramatic, emergent principles turn out to be everyday manifestations of laboratory wisdom.

Although many experimental cognitive psychologists are aware that their research may be somewhat lacking in ecological validity, they are rightly sceptical of a wholesale abandonment of experimental rigour and control in favour of a totally naturalistic approach. There are so many variables influencing behaviour in the real world, and it is so hard to manipulate them systematically, that it can become almost impossible to assess the relative importance of each variable in determining behaviour. It is hard to achieve the desirable combination of experimental rigour and ecological validity, but some of the more successful endeavours in that direction have been discussed throughout this book.

A puzzling feature of experimental cognitive psychology is the reluctance to take individual differences seriously. The typical research strategy involves using analysis of variance to assess statistically the effects of various experimental manipulations on cognitive performance, with individual differences being relegated to the error term. Cognitive psychologists who adopt this strategy seem to assume implicitly that individual differences are unimportant, and do not interact with any of the experimental manipulations.

The reality is very different. Bowers (1973) considered 11 studies in which the percentages of the variance accounted for by individual differencs, by situational factors, and by their interaction could be assessed. On average, individual differences accounted for 11.3% of the variance, the situation for 10.2%, and the interaction between individual differences and situation for 20.8%. In the light of such evidence, it seems perverse to ignore individual differences altogether!

A final problem with experimental cognitive psychology is that the emphasis has been on relatively specific theories that are applicable to only a narrow range of cognitive tasks. What has been lacking is an overarching theoretical architecture. Such an architecture would clarify the interrelationships among different components of the cognitive system. Various candidate cognitive architectures have been proposed, including Newell's (1990) SOAR architecture, and Anderson's (1993) ACT* model. Both of these cognitive architectures are based on production systems. These architectures have been applied to a wide range of tasks. However, the research community has not abandoned specific theories in favour of using these architectures, because researchers are not convinced that either of them is the "one true cognitive architecture". As long as there is no overarching, accepted architecture, experimental cognitive psychology will always suffer from a certain lack of theoretical integration.


The cognitive neuropsychological approach has become much more influential over the past 25 years. There are two main reasons for this. First, theoretical developments within cognitive psychology have helped to guide the research efforts of cognitive neuropsychologists. Second, the development of techniques such as MRI and computed tomography (CAT scans) has allowed much more precise identification of the areas of brain damage.

Cognitive psychologists often need experimental evidence to decide whether two mechanisms, structures, or processes are really separate from each other. Two examples are as follows: Are there separate short-term and long-term memory systems? Are there separate explicit and implicit memory systems? Some of the clearest evidence on such issues has come from research by cognitive neuropsychologists which has led to the discovery of double dissociations. For example, there is accumulating evidence that separate areas of the brain are involved in explicit and implicit memory (see Chapter 7).

One of the greatest strengths of cognitive neuropsychology is that it provides a good test-bed for evaluating theories originally based on the performance of normal individuals. For example, some of the most powerful evidence that there are at least two different routes involved in reading has come from studying brain-damaged patients largely lacking one route or the other (see Chapter 11).

The contribution of cognitive neuropsychology does not consist only of testing pre-existing theories. Consider, for example, the study of amnesic patients (see Chapter 7). It became increasingly clear that they could display good memory performance provided that conscious recollection of previous events was not required, and this finding was instrumental in establishing the key theoretical distinction between explicit and implicit memory. Thus, findings within cognitive neuropsychology have been used to generate new theories as well as to test existing ones.

The cognitive neuropsychological approach has been applied successfully to several areas, including perception, attention, memory, and language. However, there has probably been more research devoted to language than to any other area. Why is this? It seems that language lends itself especially well to the cognitive neuropsychological approach. The comprehension and production of language involve various skills (i.e., those of reading, listening, writing, and speaking), and cognitive neuropsychologists have shown that each of these skills has various modules or cognitive processes associated with it.

The study of brain-damaged individuals has also proved of value in establishing the functions supported by different areas of the brain. In essence, the cognitive functioning of patients having overlapping areas of brain damage is studied. A full consideration of the evidence thus obtained can reveal which part or parts of the brain are most closely associated with specific cognitive impairments.


In spite of the successes of the cognitive neuropsychological approach, it has several limitations. First, patients, even those with the same syndrome or set of symptoms, typically have rather different lesions (see Chapter 1). There is no simple answer as to what to do about the variations from one patient to another. However, the key point was expressed clearly by Knight (1998, p. 110):

People often make the mistake that a large number of patients is by definition better than a smaller one. However, as in all experimental work, the variance of the group under investigation is paramount. Your interpretations are only as reliable as the variance in your group of interest.

Second, there are often large differences among individuals having broadly similar brain damage. As Banich (1997, p. 55) pointed out, such individuals "typically vary widely in age, socioeconomic status, and educational background. Prior to brain damage, these individuals may have had diverse life experiences. Afterward, their life experiences likely vary too, depending on the type of rehabilitation they receive, their attitudes toward therapy and recovery and their social support network."

Third, as a general rule of thumb, more is likely to be learned about cognitive functioning from patients with lesions of limited extent than from those with more extensive lesions. Thus, for example, total destruction of the cortex would prevent the person affected from performing any cognitive tasks at all, but all this would tell us is that the cortex is necessary for all cognitive functions. The actual brain damage suffered by many patients is too great to permit clear inferences about cognitive functions.

Fourth, there are various important cognitive activities (e.g., creative thought; organisational planning) that seem resistant to a modular approach. For example, Fodor (1983, p. 119) argued that there was little reason to believe that advances could be made on the neuropsychology of thinking, because "in the case of central processes you get an approximation to universal connectivity, hence no stable neural architecture." However, more recently, some progress has been made (see Chapters 6, 14 and 15). For example, there have been attempts (e.g., Eslinger & Damasio, 1985; Shallice & Burgess, 1993) to apply a modular approach to the functions of the central executive of the working memory system. There is nevertheless a danger that the modular approach exaggerates the extent to which cognitive functions are localised within the brain (Farah, 1994b). As Banich (1997, p. 52) noted, "we must not forget that the brain is comprised of about 50 billion interconnected neurons. Therefore, even complex cognitive functions for which a modular description seems apt rely on a number of interconnected brain regions or systems."

Fifth, even though the study of language may be regarded as the "jewel in the crown" of cognitive neuropsychology, there are significant limitations in that research. There has been a substantial amount of work on the reading and spelling of individual words by brain-damaged patients, but rather little on larger units of language. Many important language processes are involved in reading and spelling single words, but equally there are important additional factors (e.g., contextual influence; structural themes) which are only of relevance to larger units of language (see Chapter 13). In similar fashion, cognitive neuropsychology has been uninformative about the planning processes involved in speaking and writing.

Sixth, the study of brain-damaged patients can lead to underestimates of the brain areas involved in performing any given cognitive function. The lesion method generally only permits identification of those brain areas of crucial importance to a cognitive function, but not of those that may be partially involved. Another reason is that damage to an area that is normally used to perform a cognitive function may lead the patient to develop an alternative strategy that does not rely on the damaged area. In that case, his or her cognitive performance may be normal.

Areas of the brain in which electrical stimulation was found to interfere with picture naming. The numbers represent the percentage of participants showing interference in each brain area. Reproduced with permission from Ojemann (1991). Copyright © 1991 by the Society for Neuroscience.

Positive Thinking As The Key To Success

Positive Thinking As The Key To Success

Download this Guide and Discover How To Find And Monetize on Your Expertise And Strengths. Inside this special report, you'll discover: How positive thinking is one of the key factors in a successful life. Five ways and tools to help you stay positive. Use these to help you keep on track. Case studies that'll inspire you to stick to your dreams. Plus much, much more.

Get My Free Ebook

Post a comment