Throughout the history of modern psychology different authorities would have given different answers to these and similar questions. Some would have defined psychological experiments broadly, others quite narrowly, and their criteria for defining the nature of an experiment would have varied widely. These changes in the definition of experimentation are related to changes in the material, social, and symbolic technology used in our science.

The first group of investigators whose members explicitly defined their research practice as psychological experimentation was formed at the University of Leipzig in Germany during the 1880s. Under the direction of Wilhelm Wundt, their work quickly attracted students from other countries who then set up psychological laboratories elsewhere in imitation of what they had seen at Leipzig. Wundt's laboratory thus became an early model for the first generation of experimental psychologists.

What were the major features of this model? Above all, the practice of psychological experimentation was explicitly derived from physiological experimentation. In fact, Wundt generally spoke of "physiological psychology" when referring to experimental psychology. He had qualified in medicine and was well versed in physiological research. His psychological experiments relied heavily on the material technology associated with the physiology of his time. This technology involved two categories of hardware: first, apparatus for exposing experimental subjects to controlled and precisely known forms of stimulation; second, apparatus for recording and measuring responses. Figure 1 shows how some of this [p. 234] hardware was displayed at the Chicago World Fair in 1893. This glass case display of brass instruments was the first of many public relations exercises designed to acquaint a wider American public with the scientific aspirations of the "New Psychology." Joseph Jastrow, the first American psychology PhD, organized the display for the fair.

Figure 1 Chicago World Fair, 1893. (Archives of the History of American Psychology)

The goal of this early variant of psychological experimentation was quite analogous to the goal of physiological experimentation. Where the physiologists aimed at exploring the material processes underlying the functioning of normal biological organism, the psychologists sought to illuminate the mental processes characteristic of the normal individual consciousness. "Normal" here excluded not only pathological processes, but also immature or idiosyncratic ways of functioning. "Physiological psychology" therefore did not experiment on children or on cases from the clinic, and it regarded individual and social differences among experimental subjects as a nuisance to be avoided by the appropriate selection of subjects and of research topics.

This imposed severe restrictions on the scope of the early experimental psychology. Sensory processes, usually investigated by means of psychophysical methods, constituted by far its largest research area, though perceptual processes claimed an increasing share of the experimenter's attention. Research on reaction times, though receiving much less interest, was theoretically significant from the beginning because Wundt thought that it could be used to explore central processes like attention and volition. But most of the topics that came to define 20^th-century psychology were not accessible by means of these early, quasi-physiological methods.

Figures 2 and 3 illustrate the investigation of sensory processes in the classical, Wundtian tradition. In Figure 2 the experimenter (sometimes called the "manipulator" in early studies) presents the subject (often called the "observer") with a series of different weights shielded from sight by a wooden screen. The [p. 235] observer has to compare the weights but is limited to a choice between the responses "heavier than," "lighter than," or "equal."

Figure 2 Experiment with just noticeable differences (jnds) of weights (Givler, 1920).

In Figure 3 a mental instrument is heated by passing warm water through it. It is then brought down lightly on the squares of a grid previously inscribed on the observer's hand. The observer reports sensations of warmth and touch when they occur. For many spots on the skin there may be no discrimination of temperature and touch sensations. The two upright pieces of apparatus in the foreground allow both warm and cold water to be passed through different tubes. Karl Dallenbach, a product of the Cornell laboratory established by Wundt's student Titchener, was carrying out such experiments between 1927 and 1937.

Figure 3 Touch sensation experiment (Murphy, 1935).

Several years before the founding of his Leipzig laboratory, Wilhelm Wundt had become interested in the study of reaction times. He regarded these as making possible a "mental chronometry"; that is, the timing of mental processes, especially decisions or acts of volition. He utilized methods pioneered by F. C. Donders, a Dutch physiologist. Figure 4 illustrates a simple reaction time situation. One person controls the presentation of stimuli while the other "reacts" by depressing a telegraph key. In this case, the duration of each reaction is recorded by a Hipp chrononoscope.

Figure 4 simple reaction time experiment (Givler, 1920).

But Wundt accorded much greater significance to choice reaction times, where the experimental subject must choose the proper reaction to different stimuli. This is illustrated in Figure 5. The subject (top) must discriminate between two lights which are presented by the experimenter (bottom) who, in [p. 236] this case, is located in another room of the building. The subject must then react by pushing one of two telegraph keys according to prior instructions. In this case, the duration of each reaction is recorded by a Dunlap chrononoscope (bottom right). Such choice reaction times were found to be longer than simple reaction times. As the assigned task becomes more complex, the reaction time also increases. Wundt argued that incremental increases in reaction time indicated the operation of deeper mental processes.

Even as the fame of the Leipzig style of experimentation spread, others elsewhere were laying the foundations for different kinds of experimentation. Some medically oriented investigators, especially in France, began using hypnosis for a systematic study of human reactions under abnormal conditions. These were "clinical experiments" directed at the illumination of psychopathology rather than the exploration of normal adult consciousness.

Figure 5 Complex reaction time experiment (Murphy, 1935)

Figure 6 Charcot lecturing on Blanche Wittmann. (National Library of Medicine).

The painting reproduced in Figure 6 shows Jean-martin Charcot, the most famous of the French investigators, demonstrating a case of "la grande hysterie" to a medical audience. Charcot was something of a showman, but others, including the young Alfred Binet, were subjecting hysterics to experimental procedures under laboratory conditions.

Figure 7 Wundt in his laboratory (Ca. 1910). From left to right: M. Dittrich, W. Wirth, W. Wundt, O. Klemm, and F. Sander (Bringmann & Tweney, 1980).

When we compare these French studies to what was happening in Wundt's Leipzig laboratory, we become aware of the fact that, in addition to their material technology, psychological experiments also involve a social technology. Figure 7 shows Wundt and some of his collaborators in the Leipzig laboratory. These men were Wundt's students and graduate assistants, and they alternated in the roles of experimenter and experimental subject. Whatever role they took, they were well informed about the general purpose of the experiments in which they participated. That was very different from clinical experiments in which there was typically a vast gulf, both socially and intellectually, between experimenters and subjects. A juxtaposition of the social situation in figures 6 and 7 illustrates this very well.

More generally, the social technology of psychological experimentation with human subjects involves an elaborate set of rules and relationships which provides a social framework for the collection of psychological data. Within this framework, the roles of all participants in the experimental situation are strictly defined.

Historically, the social technology of experimentation has changed as much as the material technology. In some of the earliest experiments that gave rise to long research traditions, the investigator simply experimented on himself. This was true of Fechner's psychophysical studies and of Ebbinghaus's work on memory. But as the material and symbolic technology of experiments became more complicated, a division of labor was introduced, so that one participant supervised the application of the [p. 237] technology, while the other was the source of the psychological data which the experiment was designed to obtain. Thus, a role division between experimenters and subjects came into being --though in the Leipzig model of experimentation these roles were frequently exchanged among the same set of investigators, who generally used each other as subjects.

That exchange could not happen in clinical experiments, where medical experimenters studied pathological phenomena. Here, there was a permanent division between experimenters, who controlled the experimental situation as well as the knowledge it produced, and the subjects, who were required to carry out instructions. As psychologists extended the scope of their experimental investigations to include psychologically naïve individuals, the social technology of their experiments came to resemble that of the clinical experiment rather than that of Wundtian experiment.

While the social and material technology of psychological experimentation was being established in France and Germany, another important development was taking place in England. In London, Francis Galton launched a large-scale project of "anthropometric measurement" which had far-reaching consequences for the way in which psychological experimentation evolved in the 20th century. Where Wundtian experimentation regarded individual differences as a nuisance, Galton made them the main focus of his research. In his anthropometric laboratory, large numbers of people had themselves investigated to find out how their visual acuity or strength of grip, for example, compared with that of others. This required quantitative measures of individual performance and statistical techniques for comparing individuals.

In pioneering the practical use of such techniques, Galton initiated a line of research which resulted in the transformation of psychological experimentation during the first half of the 20th century. Instead of limiting themselves to experimenting on a few subjects representing either normal or specific abnormal states of mind, psychologists could now focus on the common effects of experimental interventions on groups of individuals. This they accomplished by the statistical segregation of individual differences. As a consequence, the range of psychological topics to which the method of experimentation could be applied was enormously extended.

Eventually statistical considerations came to dominate not only the analysis of experimental data but the very design of psychological experiments. This is an example of historical change in the symbolic technology of experiments. All experiments depend on some sort of symbolic technology as much as they depend on material and social technology.

For many experimentalists the statistics of individual differences were a tool that could make the traditional search for general effects more effective. But others lost interest in this traditional goal and followed Galton in making quantified individual differences the primary object of study. First among this latter group was J. McKeen Cattell, one of many Americans to have studied under Wundt at Leipzig. There, he had participated in reaction time studies and even became Wundt's first assistant.

However, once Cattell became acquainted with Galton and Galton's methods, his interest in individual differences burgeoned. Back in America, he lost no time setting up a program for the study of such [p. 238] differences among college students. For Galton's awkward term "anthropometry" he substituted "mental tests." The material of his tests was not new, but the symbolic and social technology owed its inspiration to the pioneering efforts of Galton. The goal was no longer the investigation of the generalized human mind but, as historian E. G. Boring put it, "a description of human nature in respect of its range and variability" (Boring, 1950, p. 539).

With such a goal, the social technology of experimentation had to change too. It was no longer appropriate to limit oneself to a small number of sophisticated experimental observers as a source of psychological data. Large numbers of psychologically naïve subjects--even children--had to be studied if one wanted to explore the range and variability of human nature. That meant that psychological experimentation became an interaction among anonymous strangers.

But it was precisely this feature which opened up new horizons for psychology, for it made possible the large-scale institutional usage of psychological methods. There is considerable irony in the fact that Cattell himself was dismissed from Columbia University for his pacifist views during World War I, for it was at this juncture that the U.S. Army provided the institutional setting for the first really large-scale application of the approach that Galton and Cattell had pioneered. Well over 1.7 million recruits were subjected to mental testing.

Although this testing movement employed mostly paper and pencil methods, some aspects of the traditional experimental technology did survive within the new framework. By World War II, for instance, reaction time tasks had been incorporated into the process of selecting military personnel. Figure 8 shows complex discrimination reaction time tasks being performed by U.S. Army Air Force [p. 239] cadets. This was one of a standard battery of aptitude tests aimed at determining relative skill or ability rather than at gaining insight into what all human minds have in common.

Figure 8 Group testing of complex reaction times (U.S. Air Force).

The practical applications of psychological experimentation had first been energetically promoted by Hugo Munsterberg, who had spent some time in Wundt's laboratory in its early days and then was invited to run Harvard's psychology laboratory. He remained there until his death in 1916, increasingly active in suggesting new ways in which psychological experimentation might be applied to business, industry, and the law.

One particularly successful example of such application was the lie detector. An early version, developed by William Marston, was designed to indicate changes of blood pressure that accompanied changes in emotional states.

Leonarde Keeler built a commercially viable model of the lie detector in 1926. This "polygraph" apparatus measures changes in a suspect's heartbeat, blood pressure, respiration, and electrical skin conductance. The pens of the polygraph produce a permanent record of these measurements onto a moving strip of graph paper. The record of "relevant" questions is then compared with that of interspersed "neutral" questions. Figure 9 shows Keeler, standing in the middle as he questions a prison guard about a prison break in 1942.

Figure 9 Leonarde Keeler and his polygraph (Chicago Sun-Times).

By then, psychological experiments had acquired a certain mystique for the lay public, and excessive claims for what they could accomplish were not uncommon. This eventually led to greater awareness of the social and ethical responsibilities of experimenters.

Bibliography

Boring, E. (1950). A History of Psychology (2nd ed.). New York: Apleton-Century-Crofts.
Bringmann, W., & Tweney, R. (Eds.). (1980). Wundt Studies: A Centennial Collection. Toronto: C.J. Hogrefe.
Danziger, K. (1990). Constructing the Subject: Historical Origins of Psychological Research. New York: Cambridge University Press.
Givler, C. G. (1920/1922). Psychology: the Science of Human Behavior. New York: Harper.
Keeler, E. (1984). The Lie Detector Man: The Career and Cases of Leonarde Keeler. New York: Telshare Publishers.
Murphy, G. (1935). A Briefer General Psychology. New York: Harper.
Pearson, K. (1914/1930). The Life, Letters and Labours of Francis Galton. Cambridge, England: Cambridge University Press.