¹ Milgram, Stanley, Obedience to Authority (New York: Harper & Row, 1974)Google Scholar. Page numbers in the text refer to this work. For a much more tentative statement of these substantial conclusions see Milgram, 's ‘Some Conditions of Obedience and Disobedience to Authority’, Human Relations, 18, No. 1 (1965), especially 74–75CrossRefGoogle Scholar. This article is reprinted in Miller, A.'s useful anthology, The Social Psychology of Psychological Research (New York: The Free Press, 1972)Google Scholar, along with critical papers (see especially Chapter 2).

² Curiously the learner-confederate does not seem to ask for or receive the sample shock that is so carefully offered to the subject. In most experimental conditions the victim appears to hop into the waiting electric chair as if he were visiting his neighbourhood barber.

³ Called by Milgram, ‘Experiment z (Voice-Feedback)’, 34.Google Scholar

⁴ It should be noted that although I speak of the Milgram experiments as if there were only one model, there were in fact eighteen experimental conditions which involved variations of physical and ‘psychological’ contiguity. For a useful summary of the main variations see ‘Some Conditions’, 57–75.Google Scholar

⁵ Although a great deal that I have to say in dealing with the first category of conclusion bears on the other two, I do not directly discuss the latter kinds of inference in this paper. Aside from considerations of length I have two reasons for this omission. First, as Milgram seems to recognize (36–38), the changes in behaviour discovered in varying the nearness of the subject to the victim and/or the authority appear to be what one would expect independently of any experimentation. Second, the claims of analysis offered by Milgram do not seem to hold out the promise of the rest of his work. Witness his ‘law’ of disobedience (‘… disobedience results when net strain exceeds the strength of the binding forces’ (154)), a formulation which—contrary to common fact and some of Milgram's own results (84–85)—oddly entails that one cannot disobey with sanguinity.

⁶ ‘Some Conditions’, 59.Google Scholar

⁷ According to Milgram (13) the moral maxim the subjects in his laboratory are pitted against is: ‘one should not inflict suffering on a helpless person who is neither harmful nor threatening to oneself’. Obviously we must assume a ceteris paribus clause with very healthy range if we are to protect this principle from clear counter-examples like the doctor's knife and the dentist's drill.

⁸ Orne, Martin T. and Holland, Charles H., ‘On the Ecological Validity of Laboratory Deceptions’, International Journal of Psychiatry, 6 (1968)Google ScholarPubMed, reprinted in Miller, , op. cit., 128.Google Scholar

⁹ Ibid.

¹⁰ Milgram, , ‘Interpreting Obedience: Error and Evidence’Google Scholar, in Miller, , op. cit., 143.Google Scholar

¹¹ Milgram, , ‘Behavioral Study of Obedience’, Journal of Abnormal Psychology, 67 (1963), 375.Google ScholarPubMed

¹² Ibid.

¹³ ‘Interpreting Obedience: Error and Evidence’, 140.Google Scholar

¹⁴ Ibid.

¹⁵ Ibid., 141.

¹⁶ This point has often been made before, see e.g., Orne, and Holland, in ‘Ecological Validity’, 125.Google Scholar

¹⁷ ‘Interpreting Obedience: Error and Evidence’, 140.Google Scholar

¹⁸ Mantel, David Mark, ‘The Potential for Violence in Germany’, Journal of Social Issues, 27, No. 4 (1971), 110–111.Google Scholar

¹⁹ ‘Ecological Validity’, 128.Google Scholar

²⁰ According to Milgram some of the obedient₁ subjects do race to the end of the shock board. See e.g. 76.

²¹ And it goes without saying that stress behaviour (fits, nervous laughter) is to be expected from subjects who find themselves participating in an experiment that they perceive to be incoherent.

²² Mantel, op. cit. is equally unequivocal. He interprets Milgram's findings as showing that the average man in the street, ‘the same one who rarely comes to the aid of someone in need’, can be persuaded with case ‘to impose a series of electric shocks on an immobilized victim until he is unconscious or dead’ (102).

²³ I shall not be concerned with any general problems of generalization, but only some of those that are specific to Milgram's experiments.

²⁴ And confidence simply withers when one realizes that only 658 of the 1,000 subjects in the experiment responded to Milgram's follow-up questionnaire (see 172).

²⁵ Rosnow, Ralph L. and Rosenthal, Robert, ‘Volunteer Subjects and the Results of Opinion Change Studies’, Psychological Reports, 19 (1966), 1183–1187.CrossRefGoogle ScholarPubMed

²⁶ For a useful survey of the literature on some primary characteristics of volunteers see Rosenthal, Robert, ‘The Volunteer Subject’, Human Relations, 18 (1965), 389–406CrossRefGoogle Scholar. Also, for a clear statement of the relatively strong needs for social approval to be found in volunteers for psychological experiments see McDavid, J. W., ‘Approval-seeking Motivation and the Volunteer Subject’, Journal of Personality and Social Psychology, 2 (1965), 115–117CrossRefGoogle ScholarPubMed. See especially his warning about the ramifications of these needs in studies of social reinforcement (117).

²⁷ Rosnow, and Rosenthal, , op. cit., 1187.Google Scholar

²⁸ As Rosnow and Rosenthal note in passing, ibid., 1183.

²⁹ Adorno, T. W. et al. , The Authoritarian Personality (New York: Harper & Row, 1950)Google Scholar. For an application of the F scale see Rosen, E., ‘Differences Between Volunteers and Non-volunteers for Psychological Studies’, Journal of Applied Psychology, 35 (1951), 185–193.CrossRefGoogle Scholar

³⁰ This question is explored in detail in Hyman, Herbert H. and Sheatsley, Paul B., ‘“The Authoritarian Personality”—a Methodological Critique’Google Scholar, and Christie, Richard ‘Authoritarianism Re-examined’, both in Studies in the Scope and Method of ‘The Authoritarian Personality’ (Glencoe: The Free Press, 1954), ed. Christie, R. and Johoda, M., 50–122 and 123–196, respectively.Google Scholar

³¹ In making out this distinction of kinds of authority I use some of the terminology of Landeson, Robert F., ‘Legitimate Authority’, American Philosophical Quarterly, 9 (1972), 338–339Google Scholar. The basic distinction to which I refer is often marked off as one of ‘in authority’ and ‘an authority’, e.g. most recently, Young, Gary, ‘Authority’, Canadian Journal of Philosophy, 3 (1974), 563–565.CrossRefGoogle Scholar

³² Landeson, , op. cit., 339Google Scholar. His term for what I here call ‘expert authority’ is ‘respect authority’.

³³ Recall the military leaders in the Third Reich who followed Hitler's orders precisely, even when they knew that the commands violated basic principles of common sense and warfare.

³⁴ Milgram clearly makes this mistake: ‘Generally, authorities are felt to know more than the person they are commanding’ (141). And note as well how he chides his subjects for their failure of knowledge in accepting ‘the experimenter's definition of the situation’ (123).

³⁵ It is interesting to note that one supposedly disobedient₂ subject whom Milgram finds cause to applaud luckily can bring his background knowledge in electronics to bear and question the expertise of the experimenter in breaking off (see 51).

³⁶ The research and writing of this paper were supported by a research fellowship from the Canada Council.