Scandinavian Journal of Psychology, 1991, 32, 82-85

SHORT NOTE

Perception: The interface between cognition and the external world

HENRY MONTGOMERY Department of Psychology, University of Gothenburg, Gothenburg, Sweden

Montgomery, H. (1991). Perception: The interface between cognition and the external world. Scandina- vian Journal of Psychology, 32, 82-85

In contrast to researchers who emphasize the distinction between perception and cognition it is claimed that the perceptual system is closely intertwined with cognition and with our comprehension of the external world. This claim is supported by evidence for two characteristic features of human information processing. One of these features is the use of mental models. If mental models are assumed to be anchored in the perceptual system it will not be necessary to assume a “little man in the head” who reads off the model. Moreover, perceptually based mental models will provide a referential semantics by anchoring mentally represented concepts in the external world. The other feature is the use of perceptually based perspectives in cognition, such as actor or observer perspectives in social cognition. It is concluded that human beings consistently interact within the real world-also when they are just thinking-and that this becomes possible via perceptual processes.

Key words: Perception, cognition, mental models, perspectives.

Henry Moniogomery, Department of Psychology, Box 14158, University ox Gothenburg, S-400 20 Gorhen- burg, Sweden I

What are the roles of perception and cognition, respectively, in our understanding of the external world? A common answer to this quesjion, particularly among cognitive psycholo- gists, is that perceptual prcms$&xovide F+ut .to cognitive processes which in turn provide an understanding of the extemal world. This view underlies RBnnberg’s (1990) recent discussion of the distinction between perception and cognition. He asserts that “stimulus information is necessary but not sufficient for comprehension, or for veridical perception”. It appears that he takes for granted that cognitive processes, given a certain perceptual input, are sufficient for message comprehension.

In the present note I will claim that the perceptual system is more closely intertwined with cognition and with our comprehension of the external world than is assumed by Rijnnberg and many other cognitive psychologists. My major claim is that perceptual system is directfy involved in the understanding of the external world and not only indirectly important via cognitive processes. This claim is supported by evidence for two characteristic features of human information processing, both of which are assumed to be anchored in the perceptual system. These features are the use of mental models and of perspectives in human infona- tion processing.

Mental models

Mental model is a fashionable concept in today’s cognitive psychology and often it is not clear what exactly is meant by mental models (c.f. Carroll, 1985). In the present paper, the concept is defined as suggested by Johnson-Laird (1983, see also Montgomery, 1988). Mental . models are assumed to be analogous to a corresponding (real or hypothetical) structure in

the external world. Mental models may be manipulated in ways which correspond to

Sand J Psycho1 32 (1991)

operations that can be performed with physical objects. The results of operations on a mental model can be “read off’ and provide a solution of a given problem.

The above definition of mental model suggests that there is a close relationship between perceptual and cognitive processes. The fact that mental models may be “read off’ indicates that the use of mental models is based on a kind of internal perceptual process. This assumption is in line with the abundance of data indicating the use of perceptual codes in human information processing (e.g., Brooks, 1967; Kosslyn, 1980; Shepard, 1978). Does this view of mental models imply that besides the ordinary perceptual system there is a “little man in the head”, who perceives the mental model? The answer should be no, if we assume that the perception of mental models is performed by the same perceptual system which perceiues the external world. In other words, the manipulation of a mental model would be equivalent to activating mental structures that also are activated in “normal” perception, i.e., in the perception of the external world. Thus, in this respect perception and cognition would be inseparable. For example, seeing in the external world that John is sitting to the left of Peter and to the right of Andrew and creating a mental model of the same fact in both cases involve activation of an internal structure of the form Andrew-John-Peter. This is not to say that ordinary perception and the perception of a mental model are totally inseparable. After all, people usually can distinguish between their perception of the external world and internally generated models. My point is that in order to avoid the-little-man-in-the-head problem it is necessary to assume that the perception of a mental model involves activation of some portion of those mental structures that are activated when people perceive facts in the external world. For example, mental models may be more sketchy than is true for percepts of facts in the real world. The models may then include some but not all of the contents of ordinary perception. Mental models may also be a fictitious character and in this way deviate from ordinary perception. However, also fictitious models may consist of elements that are activated in perception of external facts. For example, a model of a unicorn (fiction) will consist of the elements horse and horn (elements corresponding to perceivable external facts). In other words, all kinds of mental models may be anchored in the perceptual system although the layout of the models may be more or less close to perceived facts in the external world.

Johnson-Laird (1983) pointed out that an important function of mental models is to provide a referential semantics by anchoring mentally represented concepts in the external world (see also Johnson-Laird, et al., 1984). Other researchers have doubted that mental models may serve such a function. Oden (1987) argued that “mental models, no matter how literal, being nevertheless mental, can no more provide a referential semantics than any other representation system” (p. 221, see also Rips, 1986). However, if we assume that the use of mental models involves the same perceptual system that gives us contact with the external world, then obviously mental models will provide extensions to the external world via those channels that are involved in ordinary perception. For example, a person may check that the model Peter-Andrew-John refers to a state in the real world by simply using her eyes to find out who is sitting to the left and to the right of Andrew. If this activity results in activation of the structure Peter-Andrew-John, then the person will know that a model consisting of this structure depicts a fact in the real world. Thus, mental models may indeed provide a referential semantics, provided that the models are linked to perceptual processes. And it is questionable whether any other types of mental representation such as semantic networks could refer to facts in the external world (Johnson-Laird, et ul., 1984).

Perception: The interface between cognition and the external world 83

Perspectives in perception and cognition

We always perceive the external world from some perspective. For example, an object could be in front of or behind another object, but it could not be both at the same time for a given

84 H. Montgomery Scand J Psyehol 32 (1991)

perceiver. We also think about the external world in ways which appear to be closely related to perspectives in perception. Perhaps the most pervasive perspective-like character- istic of human thinking is manifested in so-called actor-observer effects (Jones & Nisbett, 1971). As actors, we perceive the situation around us but as a rule we may not perceive our own behaviour. As a consequence, we tend to explain our own behaviour in terms of constraints and possibilities provided by the external situation rather than in terms of our personal dispositions. When we explain other people’s behaviour, the perspective is differ- ent. We now perceive the others behaving, but we do not perceive those features of the situation which the others think they exploit in their behaviour. As a consequence, we tend to explain other people’s behaviour in terms of personal dispositions rather than in terms of situational constraints and possibilities.

Actor-observer effects occur both when people explain what they perceive at a given moment (Storms, 1973) and when they just think about their own and others’ behaviour (Jones & Nisbett, 1972). A possible account of this body of results is that when explaining their own and others’ behaviour people draw on their perceptual experiences both when they perceive pertinent situations and when they just think about them (cf. Fiske & Taylor, 1984). More exactly, in both cases the perspective adopted in perceptual experiences may constrain a person’s explanations of hisher own and others’ behaviour.

Conclusions

The links between cognition and perception discussed here emphasize that human beings consistently interact within the real word. People are concerned with the world even when they are just thinking and I have assumed that this becomes possible via per- ceptual processes. Drawing on their perceptual systems people construct mental models which refer to real or possible states of &airs in the world. They adopt perspectives which correspond to perspectives in their perceptions of the external world. Perhaps it is not to overstate matters to characterize cognition as imbued with perception (cf. Neisser, 1976).

Cognitive psychology has been criticized for viewing man as “buired in thoughts” apparently unconcerned with the external world (Sampson, 1981). The present account suggests that this critique may be avoided by simultaneously taking perceptual and cogni- tive processes into account in research on human information processing.

Finally, it may be noted that the present account may be seen as an attempt to “cognitivize” the Gibsonian approach to perception. The key notion of the Gibsonian approach is that perception is based on information pick-up rather than on cognitive inferences (cf. Runeson, 1988). The key notion of the present account may be stated to be the cognition is based on information picked up by the perceptual system. However, in the present short note I have only been able to hint at an argumentation for this idea. Other researchers have presented a more extensive case for the role of perception in cognition (Johnson-Laird, 1983; Neisser, 1976) and the present note draws on these researchers’ work. However, much’ further research is needed in order to clarify the links between perception and cognition as well as between the Gibsonian approach and the cognitive approach to perception.

~ ACKNOWLEDGEMENT I wish to thank Anders Biel and Ulf Dahlstrand for their comments on a previous version of the paper.

Scand J Psycho1 32 (1991) Perception: The interface between cognition and the external world 85

REFERENCES Brooks, L. (1967). The suppression of visualization by reading. Quarterly Journal of Experimental

Carroll, J. M. (1985). Satisfaction conditions for mental models (Review of Mental models). Contempo-

Fiske, S . T. & Taylor, S . E. (1984). Social cognition. Reading, Mass.: Addison-Wesley. Johnson-Laird, P. N. (1983). Mental models. Cambridge: Cambridge University Press. Johnson-Laird, P. N., Herrman, D. J. & Chaf6n, R. ( 1984). Only connections: A critique of semantic

networks. Psychological Bulletin, 96, 292-3 15. Jones, E. E. & Nisbett, R. E. (1972). The actor and observer; Divergent perceptions of the causes of

behavior. In E. E. Jones, D. E. Kanouse, H. H. Kelley, R. E. Nisbett, S. Valins & B. Weiner (Eds.), Attribution: Perceiving the c u e s of behavior. Momstown, N.J.: General Learning Press.

Psychology, 19, 280-299.

rary Psychology, 30, 693-695.

Kosslyn, S . M. (1980). Imuges and mind. Cambridge, Mass.: Harvard University Press. Montgomery, H. (1988). Mental models and problem solving: Three challenges to a theory of

Neisser, U. (1976). Cognition and reality. San Francisco: Freeman. men, G. C. (1987). Concept, knowledge, and thought. Annual Rmiew of Psychology, 38, 203-227. Rips, L. J. (1986). Mental muddles. In M. Brand & R. M. Hamish (Eds), Problems in the representation

Rdnnberg, J. (1990). On the distinction between perception and cognition. Scandinavian Journul of

Runeson, S. ( 1988). The distorted room illusion, equivalent configurations, and the specificity of static

Sampson, E. E. (1981). Cognitive psychology as ideology. American Psychologist, 36, 730-743. Shepard, R. N. (1978). The mental image. American Psychologist, 33, 125-137. Storms, M. D. (1973). Videotape and the attribution process: Reversing actors’ and observers’ points of

restructuring and insight. Scandinavian Journal of Psychology, 29, 85-94.

of knowledge and belief. Tucson: Arizona University Press.

Psychology, 31, 154- 156.

optic arrays. Journals of Experimental Psychology, 14, 295-304.

view. Journal of Personality and Social Psychology, 27, 165-175.

Received 9 May 1990