William McDougall

The English-born psychologist William McDougall was the mentor of J. B. Rhine at Duke University. An autobiographical account published in 1930 in A History of Psychology in Autobiography, volume 1, edited by Carl Murchison, is available online:
https://brocku.ca/MeadProject/McDougall/..._1930.html

I thought the section describing his interest in psychical research was quite interesting. It's worth bearing in mind that the account was written before the bulk of Rhine's research at Duke. No doubt by "physical phenomena" McDougall means physical mediumistic phenomena. Perhaps things haven't changed all that much since the 1920s:

What then of my dabbling in Psychical Research? What is my apology for such "pandering to superstition"? It is probably true that the majority of those who have taken an active interest in this field have done so in the hope of providing surer foundations for religious beliefs, especially for the belief in the continuance of personality after the death of the body. I was led to make some study of this field by my desire to know the truth. Here, it seemed to me, was a body of ancient beliefs all of which Science seemed utterly to deny. Yet the ground of such denial was plainly inadequate. It was in the main an inference from the assumption that the universe is a strictly mechanical system. Here were phenomena alleged to occur in all times and places, an allegation supported by a body of strong testimony. And Science frowned upon it all and said: "Such things cannot happen." As usual I was thrown into rebellion against this orthodoxy. Further, I saw in the Society for Psychical Research a body of earnest seekers after truth, conscientiously using methods which might reveal truth; and these researches were largely in the field of psychology. Yet not only scientists in general, the philosophers, the churches, and the men in the street, stood coldly aloof or actively scoffed, but also the psychologists. And it seemed to me a scandal that psychologists should refuse to lend a hand or at least moral support to this heroic effort. Therefore, though without much hope or anticipation that any phenomenon (beyond those that fall under the head of telepathy) would he established, I threw myself to the support of Psychical Research. I felt that even a purely negative result of a long sustained cooperative research would be of great importance. For, until such research shall have been made and shown to be incapable of finding any basis of reality in the alleged supernormal phenomena, the world must continue divided into ignorant partisans and ignorant deniers.

It is peculiarly difficult to maintain a strictly scientific and impartial attitude in this sphere; a fact illustrated by the very small number of persons who have succeeded in doing so. It is difficult to avoid the influence of the confidence of the scientific world in the adequacy of its own principles, without falling under the contrary influence of traditional religion. But also a positive temptation of a very real nature besets the inquirer into these obscure questions; especially, if he has any reputation to lose or to throw into one or other scale. If, on investigating some notorious case that has excited popular interest, he hastily and roundly denounces it as purely fraudulent, he earns the applause of one half of the world ; but, as I know from my own early experiences of such sensational "exposures," he does little or nothing to clarify the field. If, on the other hand, he affirms its genuineness as an instance of supernormal happening, he wins the plaudits of the other half of the world and is accepted as a shining light among them. But if he devotes much careful study to it, and renders a judicial report, balancing carefully the pros and cons, then he becomes to both parties an object of vituperation and contempt. Although the last fate has been mine, I nevertheless find a certain satisfaction in having maintained the scientific attitude of impartial inquiry in spite of all difficulties and unpleasant consequences.

I have served on the council of the English Society for many years. I have presided over it and over the American Society, and have taken an active part in founding the new Boston Society. And, though my contacts with the field in America have brought many very disagreeable incidents, I do not repent. I have given the minimum of support which, as a psychologist occupying a position of some slight influence, I could give without reproaching myself with cowardice. If I had not found it necessary to earn some income, I should perhaps have chosen to give all my time and energy to work in this field. During my thirty years of Psychical Research I have grown rather more skeptical of the "physical phenomena" (though even now I am not prepared to assert that they do not occur) and more inclined to believe in the reality not only of telepathy but also of some of the other "mental phenomena."

Another section of McDougall's autobiographical account deals with another of his unorthodox areas of interest - Lamarckian evolution. He describes the situation when he accepted the Chair of Psychology at Harvard in 1920:

There was vacant a small equipment for animal psychology. I eagerly seized the opportunity to begin an experiment I had long contemplated. In my Cambridge days I had rebelled as usual against the then all-dominant neo-Darwinism or Weissmanism. It seemed to me that the only ground of the dogmatic rejection of the Lamarckian theory was purely a deduction from the mechanistic dogma in biology ; and I had urged that some strong scientific society should initiate and maintain, in a way not possible to any individual, a prolonged experiment designed to settle the Lamarckian question once for all, using preferably dogs as the most likely material. Now, in 1920, the question seemed just as open as in 1890, and no nearer a decision. Meanwhile, I had become more firmly convinced that the mechanistic dogma is no valid basis for biological deductions.

It was clear that, if I should use dogs as my material, I could not hope to live long enough to carry the experiment to a conclusion. I chose, therefore, the white rat which, in addition to so many other advantageous features, breeds rapidly. So, with a small group of graduate students, I set out on this fool's experiment. Yet not altogether foolishly, for, even though the issue might be entirely adverse to the Lamarckian hypothesis, a clear-cut negative issue of a well-planned and long-continued experiment would he not altogether without value; since no such experiment had been made. And, in any case, the question at issue seemed to me the most important question yet formulated by the mind of man and clearly susceptible of solution by experimental procedure. And a positive answer indisputably established by experiment would not only give us a working theory of biological evolution, but would be a heavy blow to the mechanistic biology. It would place mind at the very heart of the evolutionary process, instead of leaving it as a by-product of that process, an unintelligible excrescence upon life.

The experiment, now in its tenth year and its twenty-forth generation, seems to promise a clean-cut and indisputable proof of the reality of Lamarckian transmission. If, in the next few years, this promise should be amply realized, the work will rank as by far my most important contribution to science; although the execution of it will have required little but great confidence in my own judgment and dogged persistence. This work has absorbed all the time and energy I have had for experimental research.

McDougall trained successive generations of rats to choose one of two exits from a water-filled tank by administering electric shocks when they tried to leave by the other. He found that the number of attempts required to train the rats decreased with each of the first few generations, apparently demonstrating that acquired knowledge was being inherited by the rats' descendants.

The experiment took many years to complete, and has apparently been repeated only twice. This is what Wikipedia has to say about it:
The basic concept of inheritance of acquired characters was widely rejected in the early 20th century. However, in the 1920s, the Harvard University researcher William McDougall studied the abilities of rats to correctly solve mazes [sic]. His reports claimed that offspring of rats that had learned the maze were able to run it faster [sic]. In his data, the first rats would get it wrong 165 times before being able to run it perfectly each time, but after a few generations, it was down to 20. McDougall attributed this to some sort of Lamarckian evolutionary process. However McDougall's results have never been replicated by other experimenters, and have been criticised for having several methodological problems and poor record-keeping.
https://en.wikipedia.org/wiki/Inheritanc...cteristics

In this case, the standard sceptical comment, "McDougall's results have never been replicated," is completely wrong. They were replicated by W. E. Agar and others at the University of Melbourne. In their final report, published in 1954, they wrote:
... although an accurate comparison of the rate, and extent, of changes in learning in the two experiments cannot be made, it is clear that the improvement which characterized our first twenty-eight generations closely parallels that of McDougall's thirty-two generations and it seems probable that the same factor, or factors, operated in the two experiments.
http://jeb.biologists.org/content/jexbio...7.full.pdf

But Agar et al. found a partial reversal of the effect in subsequent generations, and more importantly they found that the changes in learning were closely paralleled in a control group of rats whose ancestors had never been trained (one of the criticisms of McDougall's experiments was that there wasn't an adequate control group). They therefore rejected the effect as evidence of the inheritance of acquired characteristics, instead suggesting that it was the result of long-term fluctuations of the general health of the rat colonies. (Whether this was really the reason I wonder, as proper records of health weren't kept, and the correlation between fertility - as a correlate of health - and training scores was too weak to be statistically significant for the control group over the whole course of the study.)

Rupert Sheldrake later came up with an alternative suggestion - the control group were benefitting from the training of other rats through morphic resonance - outlined in a chapter from his book, "Morphic Resonance"/"A New Science of Life", available here:
https://www.sheldrake.org/about-rupert-s...-resonance

This is what Wikipedia has to say about it:
The basic concept of inheritance of acquired characters was widely rejected in the early 20th century. However, in the 1920s, the Harvard University researcher William McDougall studied the abilities of rats to correctly solve mazes [sic]. His reports claimed that offspring of rats that had learned the maze were able to run it faster [sic]. In his data, the first rats would get it wrong 165 times before being able to run it perfectly each time, but after a few generations, it was down to 20. McDougall attributed this to some sort of Lamarckian evolutionary process. However McDougall's results have never been replicated by other experimenters, and have been criticised for having several methodological problems and poor record-keeping.
https://en.wikipedia.org/wiki/Inheritanc...cteristics

In this case, the standard sceptical comment, "McDougall's results have never been replicated," is completely wrong. They were replicated by W. E. Agar and others at the University of Melbourne. In their final report, published in 1954, they wrote:
... although an accurate comparison of the rate, and extent, of changes in learning in the two experiments cannot be made, it is clear that the improvement which characterized our first twenty-eight generations closely parallels that of McDougall's thirty-two generations and it seems probable that the same factor, or factors, operated in the two experiments.
http://jeb.biologists.org/content/jexbio...7.full.pdf

But Agar et al. found a partial reversal of the effect in subsequent generations, and more importantly they found that the changes in learning were closely paralleled in a control group of rats whose ancestors had never been trained (one of the criticisms of McDougall's experiments was that there wasn't an adequate control group). They therefore rejected the effect as evidence of the inheritance of acquired characteristics, instead suggesting that it was the result of long-term fluctuations of the general health of the rat colonies. (Whether this was really the reason I wonder, as proper records of health weren't kept, and the correlation between fertility - as a correlate of health - and training scores was too weak to be statistically significant for the control group over the whole course of the study.)

I have now corrected that Wikipedia account. I wonder whether anyone will object to the statement that McDougall's findings were replicated.