Social Psychology Through Experiment

George Humphrey,Michael Argyle

• Editore: Psychology Press
• Anno: 2015
• ISBN:
  9781317541370
• Edizione: 1
• Argomento: Psicologia
• Categoria: Psicologia sociale

Although widely taught to undergraduates, teachers, managers and adult students, practical work and demonstrations in social psychology were often found very difficult to carry out satisfactorily. Originally published in 1962, this book presented for the first time a series of experiments which would work in a classroom setting: some are modified versions of classical experiments, others were new. Several experiments are presented, together with discussion of their background and implications, in each of a number of central areas of social psychology.

Each topic has been covered by a different author, who has carried out research in the area in question, and is experienced in demonstrating the main experimental facts in practical class work. The editors have written a challenging introduction, in which some of the basic issues involved in experimental work in social behaviour are raised.

Chapter 1

SOCIAL PERCEPTION

HENRI TAJFEL

Oxford University

Social perception is not one of the best ordered areas of enquiry in psychology. Some of it is hardly ‘social’; some hardly ‘perceptual’. Investigations, referred to in one context or another as concerned with it, add up to a patchwork, not to a pattern. Therefore, to attempt a definition (or a series of definitions) of social perception would be a sterile task. Loosely, this blanket term stands for an agglomeration of studies attempting to find some regular and predictable relationships between man’s general orientation towards his environment, and the multitude of social factors which, in one way or another, determine this orientation or contribute to it.

These relationships may be of several kinds. For example, social influences may affect our perception of the physical environment. Conversely, changes in the physical environment may affect the way in which we react to various aspects of our social environment. As any addict of Punch or of New Yorker knows, only a precarious hairbreadth distinction separates the dignified from the ridiculous: let the rain fall, or the scene be set in a tent in the jungle, and the otherwise respectable habits of dress, speech, gesture, or courtesy become fit material for a collection of ‘funniest cartoons of the last fifty years’.

These relations between the ‘physical’ and the ‘social’ represent one only amongst the trends of interest in social perception. Our social environment consists, in the main, of people, of the ways in which they behave towards each other and towards ourselves. It is an extremely complex environment, and in order to make sense of it we must simplify, classify and label. An individual may be of interest to us either as a member of a specified group or class of people, or in his own unique capacity. To an anti-Semite, a Jew may appear as having certain attributes: the interpretation of the Jew’s appearance and actions may have its roots, entirely or predominantly, in the fact that he is known to be a Jew, and it may be to a large degree independent of the individual differences between Jews. At the other extreme, an intimate friend may be perceived or judged as being possessed of one attribute or another, independently of whether he is a Siamese, a lawyer, or a mountaineer.

It will be obvious that ‘pure cases’ at either extreme hardly exist. A set of characteristics may be attributed to someone because he happens to belong to a group about which we have preconceived ideas; obviously his own individual characteristics will also, under various conditions, become important. Conversely, an intimate friend may be seen as representative, in some respects, of his group; our ideas about this group will then serve the dual function of emphasizing some of our friend’s traits and of providing a ‘causal’ explanation for their existence.

The differences between these two forms of reaching conclusions about people are quite important. Inferences based on class membership are primarily deductive; those based on individual characteristics primarily inductive. Research on the first type of inferences must be mainly concerned with the relationships that exist between the assignment of an individual to a class, and the perceptual and evaluative consequences of this assignment. On the other hand, some of the most crucial and difficult tasks in the study of inferences from individual characteristics are in discovering what cues are selected as important and relevant, what use is made of them, and what is their objective validity.

These various areas of enquiry in social perception can be grouped in a different way along the lines of the distinction made by MacLeod (1951) between ‘perception of the social’ and ‘social factors in perception’. The points of departure of research on both sides of this slightly artificial fence were very different. The study of the ‘perception of the social’ concerned itself mainly with the ways in which people perceive other people; ‘social factors in perception’ came to mean the effects of a variety of socially derived influences on the perception of the physical environment. As Hochberg (1957) pointed out, the preoccupation of the psychologists studying these influences was with perceptual processes, and the principal object of their research to help elucidate some general problems in perception. The interests of those concerned with ‘perception of the social’ were in the fields of social behaviour and of personality, and their findings were used as a tool in the analysis of social processes, or for diagnostic purposes in clinical practice.

It must be added that this enticing symmetry is to some extent misleading. The rough correspondence of one aspect of work in social perception with perceptual interests, and of the other with social and personality interests, is by no means without exceptions. ‘Perception of the social’, or more simply, of people, was also seen by some as providing a testing ground for general perceptual theory, and in some cases, interests in perception and in personality have been fused in investigations concerned with the relationships between the two.

In a field as wide and varied as this, any choice of more or less representative experiments must be considered arbitrary and expendable: any number of other choices is possible and equally well justified. The experiments to be described below have been chosen from three areas of research indicated earlier:

(i)	The effects of social factors on perception of the physical environment;
(ii)	The effects of labels (‘stereotyping’) on perception of people;
(iii)	The nature of inference in perception of people.

SOCIAL FACTORS IN PERCEPTION

Research on the role played by social factors in the perception of physical environment concerned itself mainly with two problems: is perception of the physical attributes of an object affected by the value or the emotional significance to us of this object? And: how is our perception of the physical world affected by the information that we receive about it from various social sources?

Emotional significance and value. The results of the first group of studies have been a subject of controversy for a number of years; this controversy cannot be gone into, or even less, resolved, in these pages. It seems, however, that two statements which have sufficient empirical support can now be made. First, under some conditions, value or the emotional relevance of an object may affect the perception or judgement of this object. The size of valued objects is sometimes over-estimated, or differences in value between a series of objects may induce an exaggeration of judged differences in size (or in other physical properties) between them (McCurdy, 1956; Tajfel, 1957). Secondly, it has been shown (see for example Singer, 1956, or Dixon, 1958) that the emotional significance of a stimulus may have an effect on its recognition and on some attendant phenomena: sometimes it increases the sensitivity, sometimes it prevents swift and efficient recognition.

This must not be taken to mean that the main effect of social factors is to produce ‘distortions’ in perception. Attending, or even sometimes ‘refusing’ to attend, to some selected aspects of the physical world may be useful, and can be conceived as one facet of the general phenomenon of perceptual selectivity. One of the prime conditions of survival is the capacity to deal selectively with the infinite amount of sensory information impinging continuously on the organism, and to ‘notice’ only those aspects of the world which are relevant at the time. Similarly, some shifts in the perception or judgement of size or of other physical properties of objects can be shown to serve a function: sometimes awareness of differences (and accentuation of these differences) between objects may turn out to be more important than minutely accurate judgements of their size.

Social and cultural consensus. What we know about an object, as distinct from what we see, hear, smell, or taste of it, affects the manner in which we perceive it. One of the important and reliable sources of knowledge about the physical world is the information about it transmitted through various social channels: consensus of opinion, tradition, language, etc.

The effects of social consensus on judgements of length have been investigated in the well-known studies by Asch (1952), followed by many others. These studies, conducted in the miniature context of small groups, have shown that many people are in a state of conflict when the information provided by their senses contradicts flatly the information that they receive from others. The situation was very simple: the majority of a group, in conspiracy with the experimenter, reported judgements of length which were obviously wrong. Judgements of length obtained from the unknowing ‘victims’ in the group provided the data. Information received from other people about the simple physical properties of objects is usually reliable and fits in with the sensory information. When the two flagrantly conflict, the least that happens is doubt, hesitation, and conflict; the most–a change in judgement in the direction of the group consensus.

Asch’s experiments were conducted in an American college, and some tend to suspect these institutions of providing a favourable ground for development of strong tendencies towards conformity. However this may be, similar results have been obtained elsewhere. An unpublished experiment by Bullimore and the author conducted in an English school gave results closely agreeing with those reported by Asch. Milgram* has recently worked on the same problem (using estimates of duration of continuous tones) in Norway and in France, and found social pressure to be effective in both countries, though less so in France than in Norway.

Consensus of opinion in groups, small or large, is by no means the only social influence which comes to pattern our view of the physical world. Cultural differences in familiarity with objects and shapes are also reflected in some perceptual phenomena. At the turn of the century, Rivers (1904) has shown this to be the case in the perception of some geometrical illusions. Similar findings about an illusion the extent of which is thought to depend upon familiarity with windows and all sorts of other rectangular objects have been reported recently by Allport and Pettigrew (1957) in a study conducted in South Africa.

In summary, it can be said that when the experimental conditions are so arranged that a discrepancy exists between the two main sources of information about the physical world, the sensory and the social, this discrepancy reflects itself in a conflict. A prediction about the mode of resolution of this conflict–towards the social or towards the sensory–which would be valid in all circumstances, cannot be made. The perceptual end-product depends upon a number of attendant conditions, the most important of these being probably the clarity which with the sensory information is received.

Language. Languages differ not only in their grammar, syntax and phonetics, but also in their idiosyncratic ways of slicing the external world. Rivers (1904) found already that the Todas have ‘… a tendency to discriminate greens, blues, and violets less definitely than reds and yellows, and that the deficiencies in nomenclature for the former group of colours are accompanied by a certain degree of deficiency in their discrimination’ (p. 328). He thought, however, that this was adding to the evidence ‘… in favour of the view that one of the factors upon which the defective colour nomenclature of the lower races depends is some degree of defect in sensitiveness’ (p. 392).

Different languages tend to impose different groupings on such physical ‘continua’ as the range of colours. These groupings can be described in terms of three principal characteristics: the number of labels assigned by a language to a continuum; the regions on the continuum where the assignment of one label tends to shade over into the assignment of another; and the consistency with which a particular label is assigned to a particular stimulus on the continuum. Lenneberg and Roberts (1956) provide an example of the first in comparing the number of temperature terms used in Portuguese and in English. Three terms exist in English: hot, warm, and cold. In Portuguese, there are only two: quente and frio, with the intermediate terms (tepido and caldo) not frequently used. Similarly, the French term tiède is more rarely encountered than chaud and froid. Differences in the number of terms used imply differences as to the regions of the continuum to which these terms are applied. For example, Lenneberg and Roberts found that the Zuni Indians made one category of the colour region divided by th...