When I was an undergraduate student studying biology, a botany professor shared with us an article from Science magazine published in 1890. The paper was presented to the Society of Western Naturalists by its president T. C. Chamberlin. It was very influential to me in 1985, but during a move some years later, I lost my copy, so for most of the 40 years since I first read it, I have been without it. I did make several attempts at finding it, but I had recalled it was published in Scientific American, so most of my attempts had failed.
I had lost most hope of finding it and had moved onto other things, when the article came up in conversation and someone suggested using AI to find it. Sure enough, my first search identified the article and within a few minutes I had the PDF available from JSTOR saved in a drive.
I was not wrong in remembering the importance of the article in my thinking, and I see some of what I have said for years in the article. Of course, I do not believe the article itself directed my thinking, but the botany professor was on the faculty of the university that promoted such thinking.
The title of the article is “The Method of Multiple Working Hypotheses.” Chamberlin begins the article by identifying two types of thinking. One “follow[s] by close imitation the processes of previous thinkers.” The other requires one to “think independently, or at least individually, in the endeavor to discover new truth.” These types of thinking can be directed by three types of learning:
- The ruling theory—in which we follow the previous thinkers.
- The working hypothesis—in which we test an idea to discover if it is true (or at least as close to true as science allows).
- Multiple working hypotheses—in which we try to explain not only our preferred hypothesis, but as many others as we can conjecture.
Chamberlin argues that when we attempt to answer the questions “How came this so?” we tend to come up with a hypothesis that we think best explains whatever we are trying to explain. We all learned this is part of “the scientific method” and even in 1890, Chamberlin was telling us that method which was commonly taught was not how science really happens. As we are trying to make observations to determine if the hypothesis is supported by the evidence, we will be biased to believe our “intellectual offspring” and in today’s vocabulary, Chamberlin claims we will cherry-pick data to support our hypothesis and ignore data supporting other hypotheses.
The method that we explore “every rational explanation of new phenomena” is especially useful when we are studying complex phenomena. Chamberlin was a geologist, so he used the origins of the Great Lakes as an example.
Chamberlin does suggest there can be problems with the method. When we communicate about our ideas, it is difficult to describe more than one at a time. Also, the method can be challenging for young students (young here means those uninitiated, they can include adults who are unfamiliar with the field). I am also a fan of Stephen Jay Gould. In his monthly essay in Natural History and the books that collected them, he claimed he always felt his lay readers could understand the most sophisticated topics.
About halfway through the article, Chamberlin turns his focus to education. He writes, “pedagogical inquiry in the past has largely concerned itself with the inquiry ‘What is the best method?’ rather than the inquiry ‘What are the special values of different methods, and what are their special advantageous applications in the varied work of instruction?’”
Teaching is a complex endeavor, perhaps more complex than the reasons for the Great Lakes. By having multiple hypotheses, the teacher “shall have in mind the full array of possible conditions and states of mind which may be presented.” Basically, when educators understand that there may be a greater set of reasons why their teaching works (or doesn’t) they are more likely to find methods that do work.
By recognizing multiple hypotheses, Chamberlin suggests “the outcome, therefore, is better and truer observations and more righteous interpretations.” For me, it is important when deciding things like the origins of the Great Lakes that we have good observations, but it is even more important for teaching.