To read Part 1 of this series, click here.
To read Part 2.0 of this series, click here.
More on the demarcation problem…
I just wanted to make another comment on the demarcation of science and non-science before moving on. I work in a field (Experimental Psychology) that is often criticized for not being a true “science.” In fact, one of the greatest scientific thinkers of our time, Richard Feynman, dismisses the social sciences (see full video here) by saying, “Social science is an example of a science that is not a science.”
Now, remember that my argument for what makes a question or theory scientific is the degree to which it is falsifiable. Can we prove it wrong through observation, measurement, and experimentation? Furthermore, can these tests be replicated by other researchers to get the same, or different, results? According to these criteria, the demarcation between science and non-science is rather simple.
Feynman follows his comment about the social sciences with “…they don’t get any laws, they haven’t found out anything.” This, I believe, is a very misleading comment and (although I love the man) I’m rather disappointed that he made such a comment.
I think where the problem lies is how some choose to demarcate science from non-science. Instead of using the criteria I’ve outlined above, some scientists use “causal inference” as their reference point for what makes something scientific. Basically, causal inference is when we draw conclusions about a causal relationship (A causes B) where an effect (A) occurs. If we determine that this relationship is universal and fundamentally true through empirical evidence, then we can define a law. For example, the observation of an apple falling from a tree leads to the question “why does it fall?” This question leads to the hypothesis, “because there is a natural force pulling it down.” Then after decades of experimentation and peer-review we discover the laws of gravitation and relativity. These laws, in turn, explain why the apple falls to the ground, i.e., we can infer that the basic law causes the movement of the apple. For some scientists, the degree to which a field can explain phenomena through cause-effect relationships determines how scientific it is.
But what fundamental, universal laws govern human behaviour, rational decision-making, and social interactions? Are we to believe that since these phenomena are incredibly complex that we can’t study them through scientific methods? We can’t ask questions and develop testable theories?
Another comment that Feynman makes in the same interview regarding the social sciences is, “they haven’t got anywhere, yet. Maybe some day they will…” Does this imply that we should dismiss an area of inquiry as unscientific because it hasn’t “yet” contributed any laws to explain phenomena in the universe? Do we dismiss the efforts of the early natural philosophers and alchemists as pseudoscience simply because they had not yet discovered the Law of Thermodynamics or Henry’s Law? It seems to me that these early scientists were discovering associations between phenomena, which we could later explain with higher degrees of precision. Similarly, the social sciences are still at an early stage of development where they’re mostly describing associations between things.
In summary, I would counter Feynman’s comments by stating that what he’s talking about is the difference between experimental and non-experimental designs. I won’t get into describing the difference between the two, but basically the former allows for greater causal inference whereas the latter does not, and the social sciences employ a lot of non-experimental approaches.
Sorry, Feynman – we have to start somewhere.
To be continued…