Orientation Observation In-depth interviews Document analysis and semiology Conversation and discourse analysis Secondary Data Surveys Experiments Ethics Research outcomes



Social Research Glossary

About Researching the Real World



© Lee Harvey 2012–2017

Page updated 29 May, 2017

Citation reference: Harvey, L., 2012–2017, Researching the Real World, available at
All rights belong to author.


A Guide to Methodology

9. Experiments

9.1 The classic experiment
9.2 Field experiments
9.3 Ethnomethodological experiments
9.4 Summary and conclusion

Activity 9.3

9.3 Ethnomethodological experiments
Ethnomethodologists have used experiments to challenge people’s taken-for-granted views. Ethnomethodologists argue that people have shared assumptions and sets of taken-for-granteds that they use in making sense of the world (see Section

To test this out, they have undertaken experiments designed to disrupt normal sets of taken-for-granteds. These experiments, which include such things as bargaining for a fixed price item in a chain store, are called breaching experiments as they breach shared assumptions.

The aims of the breaching experiments were to show that:

1. people have shared assumptions that allow them to interact with other people as both share the same taken-for-granted ideas;

2. these taken-for-granteds were used by people to control their world;

3. people became frustrated when the taken-for-granted rules did not work;

4. people had a different set of taken-for-granteds which they used in different situations.

This is illustrated by Harold Garfinkel (1967), who asked his students to spend between 15 and 60 minutes acting out the role of a boarder in their home. They were instructed to be polite and circumspect, to avoid getting personal, to use formal address and to speak only when spoken to. Nine out of forty-nine of his students were unable to carry out the experiment for various reasons. The results from the other students varied. Some families reacted in astonishment, bewilderment, shock, embarrassment or anger, or demanded explanations: ‘What’s the matter?’, ‘What’s gotten into you?’, ‘Are you sick? and so on. Normality was restored when students explained what they had been doing. Most of the families were not amused and most had not found the situation instructive. After the explanation one student’s sister coldly replied on behalf of her family:

Please, no more of these experiments. We’re not rats, you know. (Garfinkel, 1967, p. 49)

Garfinkel used this experiment to produce situations in which the normal rules of everyday life were changed. The students did not act as they usually did and normal interaction became very difficult. The students’ families were no longer able to make sense of the situations by making use of the taken-for-granted assumptions they had previously used.

Garfinkel was also interested in the taken-for-granteds that guided conversation. As part of his ‘breaching’ experiments he instructed his students to start a conversation with a friend and (without saying they were involved in an experiment) to insist that the person clarify the sense of his/her ‘commonplace remarks’. In other words, to breach the taken-for-granted’s of everyday conversational expectations. The cases in CASE STUDY conversational experiment illustrate the approach.

Activity 9.3
Bearing in mind that some people may get upset and that there are ethical problems with using people without their knowledge in experiments (see Section 10), replicate, if possible, Garfinkel’s conversational experiment and record your results.
Did your subject(s) show similar reactions to those in the CASE STUDY conversational experiment?
WARNING Be careful not to push the respondent too far, especially if you value the friendship. If you have any reservations about this approach do not do this activity.


9.3.1 Epistemology of ethnomethodological experiments
Garfinkel (1967) was not concerned with explanations of social phenomena. Instead he was trying to reveal the limitations of sociology. For him, ethnomethodological experiments were less concerned with proving a causal connection than with revealing what we take for granted. Strictly speaking, breaching experiments are not experiments in the classic sense because they are not concerned with issues of control nor with causal explanation (see Section 9.1.4).

Ethnomethodologists argued that sociology starts out by presuming the nature of social interaction. Sociology ignores the process by which social interaction is made possible. What ethnomethodologists attempt to do is to analyse the way in which meanings are shared. They started by asking if people had shared meanings. The ethnomethodological experiments, by deliberately breaching taken-for-granteds, indicated that interaction was dependent on shared meanings. (See Section 6.1 on discourse analysis for more detail and examples of the importance of shared meanings in conversation).

People made sense of the world through these shared meanings and were often unable to cope when they were remove. Sometimes, however, when one set of shared assumptions seemed useless, subjects adopted a different set to help them make sense of the situation. Garfinkel suggested that far from being ‘cultural dopes’, people are extremely sophisticated. The experiments showed that we are quite capable of using one set of rationalities in one situation and another set in a different situation. For example, the set of taken-for-granteds that a teenager adopts with friends in a discotheque is not the same set of taken-for-granteds that he or she adopts when in a job interview.

Thus ethnomethodological experiments are not attempting to make causal connections. They are contrived situations that are designed to challenge taken-for-granteds to see how people make sense of them. Thus, the problem of control, so central to the classical experiment, is not important. Ethnomethodologists are guided by phenomenology and are interested in meanings. Ethnomethodological experiments are a tool to explore meanings and are quite distinct from positivist experiments designed to prove causal relationships.



Next 10 Ethics