The linear model implies independence between stages from input to output and involves bottom-up processing were the processing of inputs is not influenced by higher types of knowledge. The interactive model by contrast suggests that the different types of knowledge link together and involve top-down processing were higher knowledge influences the processing of inputs. Van Dijk & Kintsch (1983) proposed that people only recall the main gist when reading long text and cannot remember the actual wording. The recalled gist varying between different people as they apply their individual beliefs and attitudes surrounding the read topic. When reading we retain information of major events and themes and leave out minor details, which we feel are unimportant. We are therefore highly selective in this process and what we retain maybe linked to what is vivid or appropriate to us at that time. One explanation maybe is that, mental representations are sought and used to construct understanding. It is doubtful whether such representations are readily available without information processing. Take a person who stammers for example, they may choose to replace a ‘stammering’ word or may instead struggle on the ‘stammering’ word. During this process many mental representations may come into play, for example of words, of meanings, of the situation, of their struggle. Each thought and subsequent representation would involve processing and interaction including links to past experiences. Without information processing and representations, language would just be ‘here and now’. Under reading and understanding comprehension, Kintsch (1988, 1994) proposed the construction-integration model, were reading activation processes are used to select propositions for text representation. Kintsch proposed that clusters of highly interconnected propositions attract most activation and are then most likely to be included in text representation. Things that belong together contextually are stored in episodic text memory with construction processes involved in linking relationships.
Language understanding involves inference based on prior knowledge and the processes linking different representations such as general knowledge, lexicon, syntactic and semantic rules. Indeed Humphrey’s et al (1988) proposed a cascade model of interaction were information flows through the system as objects are identified. The model sees representations and information processing linking together to produce object recognition.
It is now necessary to consider models within the arena of problem solving.
Problem solving offers examples of how people analyse problems and the different representations and techniques used. Simon (1978) highlighted the importance of transformation problems such as used in the ‘Towers of Hanoi” problem. Simon added that a greater understanding of problem-solving has been discovered through analyzing behaviours of people solving transformation problems. Research has demonstrated that the varying ways a solver constructs a mental representation contributes to the varying ease or difficulty of solving a problem. Transfer of learning is another aspect of problem solving in which research has demonstrated positive and negative transfer between tasks. Simon (1978) adopted an information-processing view and saw problem solving success through using mental problem space. In more complicated or difficult problem solving, the solver often has to plan and work through scenarios in their mind. During this technique the working memory is heavily used. Memory is also required to search for solutions using previous knowledge and experience. This suggests that representations are brought into play were previous knowledge and experience helps construct solutions. For example schemas may be derived from past experiences of similar problem types and using these and adapting them helps produce new solutions to problems.
People try to relate what they understand about a particular problem to specific knowledge they hold about problems of a similar type. In order to solve a problem the solver goes through different mental knowledge states (Simon & Simon 1978), and processing involves different mental operations. Such mental operations will be unique between solvers based on their past knowledge and their different mental representations of the problem. Models therefore have to allow for individual differences between problem solvers. Simon (1974) also considers the role of memory, including capacity of working memory and time differences between storage and retrieval. Retrieval from memory is triggered by differing means but involves the accessibility of information.
Categorization of objects can allow objects to be represented in many different ways. Our knowledge helps us to make classifications based on certain attributes. The categorization allows things such as objects and animals to be represented in memory in varying forms, which are designed to aid recall and understanding. Therefore an object or animal is recognised from previously stored details, which include concepts and categories at varying levels or hierarchies, all linked back in some way to the original. Collins and Quillian (1969) proposed that people’s internal, mentally stored representations take the form of hierarchy. Broad categories as nodes being at the top and broken down into more specific categories as the branches descend. Each level links back and each contains defining properties. This model helps identify that processing is required to help establish links and for example distinguish between a canary and an ostrich. Without information processing of objects and attributes together with the associated mental representations the retrieved information would be limited, and less effective.
Understanding the context in which things are presented to us, is an aspect which links mental representations and processing. Our capacity to see and recognize objects is important but without information processing the context and meaning could be lost in translation and the nature of the task wrongly interpreted.
The distinction between episodic and semantic memory knowledge has a bearing on retrieval methods and successes. Studies for example have shown that an eye-witness remembers more after returning to the scene of a crime, implying that context representations were stored. Studies have also demonstrated that knowledge is organized and that deeper processing produces more effective coding. Eye-witness testimony and formal identification by witnesses would produce a link between the processing of information and the associated representations the memory of the crime may hold for the witness.
Anderson’s ACT* (adaptive control of thought) Model (1996) extends the multi-store model in terms of connecting processes of storage and retrieval as information is transferred between working and declarative memory. Declarative memory contains knowledge stored by means of a variety of different kinds of representations or schemas. Different processing types are involved in the operation of the ACT* system, including encoding used to create output representations of the external world. The framework of the ACT* system indicates processing and encoding which interlink for storage, retrieval and modification of memory. The model incorporates both mental representations and information processing, where processing is goal directed through encoding and pattern matching.
The parallel distributed processing approach looks at the micro-structure of the brain (Anderson et al, 1977). PDP models hold properties such as schema-based learning in human cognition. Such models interest Psychologists in areas of human cognition because of the natural inevitability were parallel distributed processing and representation are results of such an information processing system. The PDP approach can be used as a general system for modelling and simulating human cognitive behaviour.
Kosslyn (1975, 1976, 1980) investigated processes and representations involved in imagery. Different animals of varying sizes were imaged. Participants were asked to imagine two animals eg. rabbit and elephant and then asked to ‘see’ a feature of the smaller animal eg rabbits nose. Participants response times were longer for ‘seeing’ features of the smaller animal. The results were explained due to the larger animal filling more of the imagery space. Such findings help demonstrate that processing of information and representations co-exist.
Object recognition leads us into face recognition. An important difference is that objects recognition involves between-category discrimination whereas face recognition involves discrimination within the category of faces.
Representation of faces include varying surface characteristics, including texture and pigmentation, and processing can be holistic or based on analysis of features. Therefore demonstrating decisions being made at different levels of conceptual hierarchy. It is likely that face recognition is based on analysis of image properties. Young et al (1985) proposed a model of functional components involved in personal identification. The model highlighted representational processing were stored representations of known faces produced recognition units of measurement. The sequence of stages made up full recognition and involved processing of different features.
Humphrey’s, Lamote and Lloyd-Jones (1995) produced an activation and competition model of object recognition. According to the model, the structural description of objects visually similar to the object actually presented, are activated to some extent. An assumption made is that living things are typically more similar to other members of the same category – unlike with non-living things. According to the model, living things should generally be categorised more quickly but named slower – when compared to non-living things. The theory being that living things are more visually similar to each other. This causes more activation of irrelevant representations and name representations, both of which inhibit naming living things. The model was further developed for similar identifications to provide accounts for object recognition, and is a good example of a detailed process model.
Although not a model in it’s own right, The Stroop phenomenon is worthy of consideration when discussing representation and information processing. Stroop (1935) discovered that semantic information can interfere with a persons performance on a task. Names of colours were printed in inks of different colours, eg. The word BLUE written in RED ink. The task was to name the colour of the ink but interference had a significant effect in this demonstration of data driven automation. Colours may bring representations to the forefront of the human mind but automatic processing becomes the dominant force.
Other examples demonstrating the close link between representation and information processing, include “tip of the tongue” phenomena – where representations can be present but access to the information is a problem. People try to process and gain access to the information but for one reason or another are unable to fully complete the ‘chain’.
If the human mind was able to access and understand it’s internal representations but had no mechanism for being able to interpret or translate into known functions or correct perspectives then the result would be confusion and inappropriate behaviour. This, as many cognitive models imply, suggests a strong link between representations and information processing. There is little point in storing representations if our access to them is limited or inappropriate.
Within the space of problem solving we have seen that mental representations of problem structures and the appropriate procedures should interact to solve problems, and solutions to new problems require appropriate operators in order to progress to the goal state.
It is also evident that our problem-solving schemas are derived from experiences with similar types of problems and the processing is dependent upon the retrieval of such schemas and then adopting them to help produce new solutions. Different strategies used in attempting to solve problems rely on different processes, and different sequences of operations, in order to gain access to the stored knowledge and concepts necessary to solving a particular problem.
The information-processing approach to cognitive psychology underlies computer modeling, which has in turn helped researchers to use parallels between artificial and natural intelligence and has led to increased understanding of human cognitive processes. Although this example is not directly linked to representation it does however show the importance of information-processing and why models cannot ignore it.
In conclusion, in the absence of a particular object we rely on our representations to bring the object to the forefront of our minds. Certain models of human cognition would have us believe that representations and information processing do not combine to aid cognition. This essay however has taken an opposing stance and has tried to demonstrate that mental representations and information processing work together and that no model of human cognition is complete without the two.
References
ANDERSON, J.R. (1996). ACT: A simple theory of cognition. American Psychologist, 51, 355-365
ANDERSON, J.R., SILVERSTEIN, J., RITZ, S. and JONES, R. (1977) ‘Distinctive features, categorical perception, and probability learning: some applications of a neural model’, Psychological Review, 84, pp. 413-51.
COHEN,G., KISS, G. and LE VOI, M. (1993) Memory: Current Issues, Open University Press in association with The Open University (Open guides to Psychology series).
COLLINS, A. and QUILLIAN, M.R. (1969) ‘Retrieval time from semantic memory’, Journal of Verbal Learning and Verbal Behaviour, 8, pp.240-7
EYSENCK, M.W. & KEANE, M.T. (2001), Cognitive Psychology, A Student’s Handbook,
4th Edition, Psychology Press
EYSENCK, M. (2001), ‘Principles of Psychology’, 2nd Edition, Psychology Press
GREENE, J. (1995) Language Understanding: A Cognitive Approach,, Open University Press in association with The Open University (Open guides to Psychology series).
HUMPHREYS, G.W., RIDDOCH, M.J. and QUINLAN, P.T. (1988) ‘Cascade processes in picture identification’, Cognitive Neuropsychology, 1900, 5(1), pp67-103
HUMPHREYS, G.W., LAMOTE, C., and LLOYD-JONES, T.J. (1995). An interactive activation approach to object processing. Effects of structural similarity, name frequency and task in normality and pathology. Memory, 3, 535-586
KAHNEY, H. (1993) Problem Solving: Current Issues, Open University Press in association with The Open University (Open guides to Psychology series).
KINTSCH, W. (1974) The representation of meaning in memory, Hillsdale, N.J.: Lawrence Erlbaum Associates.
KINTSCH, W. (1988). The role of knowledge in discourse comprehension: A construction-integration model. Psychological Review, 95, 163-182
KINTSCH, W. (1994). The psychology of discourse processing. In M.A. Gernsbacher (Ed.), Handbook of psycholinguistics. London: Academic Press.
References continued
KOSSLYN, S.M. (1975) Information representation in visual images. Cognitive Psychology, 7, 341-974
KOSSLYN, S.M. (1976), Can imagery be distinguished from other forms of internal representation?: Evidence from studies of information retrieval time. Memory & Cognition, 4, 291-297.
KOSSLYN, S.M. (1980). Image and mind. Cambridge, MA: Harvard University Press.
ROTH, I. And BRUCE, V. (1995) Perception and Representation: Current Issues , Open University Press in association with The Open University (Open guides to Psychology series).
SIMON, D.P. AND SIMON, H.A. (1978) ‘Individual differences in solving physics problems’, in R.S. Seigler (ed.) Children’s Thinking: What develops?. Lawrence Erlbaum Associates.
SIMON, H. (1974) ‘How big is a chunk?’ , Science, 183, pp.482-8.
STROOP, J.R. (1935) Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643-662.
VAN DIJK, T.A. AND KINTSCH, W. (1983) Strategies of Discourse Comprehension, Academic Press
YOUNG, A.W., HAY, D.C. and ELLIS, A.W. (1985) ‘The faces that launched a thousand slips: everyday difficulties and errors in recognizing people’ British Journal of Psychology, 76, pp.495-523