The awlfulness of Milgram

[[[[[[[ FROM CHUCK TUCKER 931201 ]]]]]]]

        The "awfulness" of the Migram studies (WTP 931123.0830 MST)

        You ask some very important questions which unfortunately can
        not be adequately answered from the reports by Milgram of his
        studies. His studies major flaw (which is the case for most
        research except some done by PCTers) is that there is no data
        on individuals as individuals; no data on specimens. Ernie and
        Clark asked Milgram for these data about 25 years ago but he
        refused to provide them. The closest he comes is to provide
        some reports from a follow-up questionnaire and some post-experi-
        mental interviews with some subjects. This is why I contend that
        it would be important to replicate his studies using simulation
        and "role playing" procedures and obtain observations of
        individuals. My friend Bob Stewart disagrees with such a plan
        because he believes that the Milgram studies were unnecessary
        in the first place; observations of man's inhumanity to man,
        he contends, can be seen a million times each day as people
        attempt to coerce each other to behave in particular ways. He
        is correct but I would like to know more about such acts so,
        at the minimum, I can figure out how to resist them. But back
        to your questions.

        Milgram supplied some information related to your questions
        in an answer to critiques by Orne and Holland (reprinted in
        A. Miller (Ed.) 1972. THE SOCIAL PSYCHOLOGY OF PSYCHOLOGICAL
        RESEARCH. NY: Free Press, pp. 124-138). On any of your
        questions which you ask about "all the subjects" the answer
        is either "no" or "we don't have the information." But on
        the question "Did the subjects believe they were shocking
        another person?" Milgram provides two pieces of data. He
        asked the subjects on a question using a 14 point scale
        answer (circle the number type) "How painful were the shocks
        you last administered to him?" The answers ranged from a
        mean (no other statistical calcuations) of 13.42 (out of 14)
        to 11.7 and there seems to be no remarkable differences from
        study to study or whether subjects pulled the 450 volt lever
        or stopped and exited the study before pulling that lever
        (1972:125).

        Then there is a follow-up questionnaire (which amazingly
        appears to have a return rate of 75% !) upon which he found
        that only 4% of the subjects were certain that shocks were
        NOT being delivered. But only 56% of the subjects who
        answered the questionnaire agreed with the statement "I
        fully believe the learner was getting painful shocks" thus
        another 40% agreed with the statements that it was "probable,"
        "wasn't sure," or "had some doubts." There seems to be no
        differences (again) between those who answer these questions
        in the variety of ways offered to them and whether they
        completed the study or exited before pulling the 450 volt
        lever. Of course, there is no analysis of specimens.
        Also, in spite of the lack of good data (from my point of
        view), Milgram concludes that the subjects accepted the
        reality of the experiment. I wish that he would have
        done an analysis on those who accepted the "same" view of
        the experiments but Milgram decided not to use that
        procedure (1972: 127) [I do have some questions about
        its use in "diffused status" studies done by sociologists
        at Stanford but I think it would have been very useful
        in the Milgram studies].

        In this article by Milgram he does lean in the direction of
        an interpretation of his studies which could be useful to PCT
        and gives some infromation about competing goals which were
        in conflict in these studies. Milgram notes:

            ... there is an important developmental aspect to
            the experiment which comes to constrain and control
            the subjects behavior. The early stages of the
            experiment are quite proper, even uneventful; it
            is only gradually as the shock levels intensify
            that conflict arises. The earlier parts of the
            experiment, in which any reasonable person would
            participate, only gradually eases the subject into
            conflict; when conflict arises the subject has
            already routinized his behavior, committed himself
            to the procedures and, in consequence, is locked
            into the situation (1972:129).

        Now, Milgram does not mention it here but he had one study
        where the subjects were not told to increase the shock level
        after each mistake made by the Learner. In that study
        ("Free to choose shock level" 1974: 70-72, 60 Table 3, Column 7)
        one 1 subject (YES ONE) pulled the 450 volt lever and not one
        of the other subjects pulled a 165 volt lever. About this
        result Milgram states: "Whatever leads to shocking the victim
        at the highest level cannot be explained by autonomously
        generated aggression but needs to be explained by the
        transformation of behavior that comes about through obedience
        to orders (1974)." I disagree! All the studies have
        "obedience to orders" but this was the only study that
        did not require the subjects to increase the shock to the
        next highest level after a Learner mistake. It shocks me
        how Milgram could miss this procedure in his own study.

        But on leaning toward a PCT interpretation Milgram states:

            The parties are embedded in a social defined
            hierarchical structure and the fact dominates
            their behavior. Social sturcture is not a
            mysterious thing. From the standpoint of the
            participating subjects it is the conviction
            that another person, by virtue of his status,
            has the right to prescribe behavior for him
            (1972:130)

        and

            In the experiment, the act of shocking the
            victim is coordinated ot a set of rational
            purposes concerning advance of knowledge
            (1972: 130).

        and

            In the obedience experiments, the act of
            shocking the victim is tighltly embedded
            in a set of socially constructive purposes,
            namely, the furtherance of knowledge in regard
            to memory and learning processes (1972: 131).

        and

            ... we are not dealing with personal power of the
            experimenter as in the case of hyposis (Danny Miller
            may disagree with this judgement) but, quite explicitly,
            with the consequence of social structure for action.
            ... the purpose which authority defines are not
            senseless and stupid (as in the nitric acid study)
            but are readily accepted by the subject as worthwhile.
            ... the experiment has an important temporal aspect to
            it. It begins with mutual concent of all parties
            [Milgram does not mention that this consent is based
            on deception - CWT] and only gradually leads into
            conflict (1972: 131).

        So, Milgram at least begins to try to make some sense out of
        this work by discussing relationships. But he does not even
        marginally discuss what the conflict might be except in a
        useless (for me) discussion of psychological strain. I would
        contend that the subjects were in conflict about several goals
        and agreements.

        They agreed to take part in a SCIENTIFIC study to gain more
        information (to serve science) about how people learn and
        acquire knowledge. They agreed with the idea that reward
        and punishment had something to do with learning (sound
        familiar?). Each was randomly selected (through deception)
        as a Teacher and believed that they could have been the
        Learner just as easily. They believed in those experiments
        where there were "co-teachers" that they were also select
        by "chance." They were told that the shock would not be
        harmful to the Learner and believed it. They were all
        adults from the community not, college students getting
        money and/or course credit for doing the research (they
        were given $4.50 when they arrived at the laboratory and
        told they did not have to participate to keep the money).

        What we have in most of these situations is a triadic rela-
        tionship set (see Dan Miller. 1986. "Milgram Redux ..."
        STUDIES IN SYMBOLIC INTERACTION V7 Part A pp. 77-105. JAI
        Press for an excellent and insightful discussion of such
        relationships and their implications for the Milgram studies
        and social life). When the Learner began to protest (an act
        designed as a disturbance) the Teacher started to have a
        conflict between various goals: doing science, contributing
        to knowledge, continuing the relationship w/ the experimenter
        and/or co-teachers, being a human being and doing the experi-
        mental procedures as specified and perhaps even some "intrinsic
        control variable" evidenced by sweating, screaming, studdering
        and shaking. The experimenter (and co-teachers) in most
        studies insisted that the Teacher continue ("You must go on,"
        "You must continue the experiment," "Science demands that
         you continue," etc.). Now, we don't know in the detail
        that we require (in fact we don't even know loosely) what
        happened for each subject but those that went to the end
        either did so to accomplish the above stated goals while
        those who quit (Milgram reports some of their statements)
        because of science, knowledge, the relationship, the comitt-
        ment, the experiment was less important than the life of
        the Learner. Some believed that the Learner was not being
        hurt by the shocks as they were told by the experimenter.

        There is one more idea as to "why" some subjects did not go
        to 450 volts (this is Bob Stewart's idea): the experimenter
        "lost courage" and exceeded to the expression of pain and
        anguish displayed by the Teacher and did not insist that the
        Teacher continue on and pull the 450 volt lever.

        It would certainly be nice someday to find out if these
        speculations and this "re-analysis" (which, by the way,
        I will share the responsibility and praise for the results
        with McPhail, Stewart, Rigney and all of the members of
        CSG-L especially the co-founders of Control Theory) could
        find some evidential support from specimens in a series of
        studies explicity based on HPCT. Any suggestions?

        More than you ever wanted to know Chuck.