Astronomy is another scientific field in which valuable knowledge has been achieved with non-experimental observation. I believe that the same is true in my own area, psychology, specifically personality assessment and measurement, which relies more on correlational than experimental methods. In 1957, Lee Cronbach wrote an article, "The Two Disciplines of Scientific Psychology", in which he discusses the experimental and correlational methods, in which he encouraged combining the two approaches. His 1975 follow-up, "Beyond the Two Disciplines of Scientific Psychology," was pessimistic in tone, concluding that higher-order interactions among variables makes in unlikely that we will be able to make generalizations. Although Cronbach suggested that both the experimental and correlational approaches were equally important, I think he ignores the common perception that experimental science is superior to observational/correlational science (a perception with which I disagree).
Interesting post. It sounds like you’re talking about the perceived superiority of hard sciences over soft sciences, which I think is a useless concept because it ignores the permeable boundaries between the two. I’ve often heard the notion that personality assessment is “junk” science, and I think that’s because it is TOO generalized and people can always find exceptions to the theory. How do you overcome this conundrum?
I agree that the boundaries between hard and soft sciences are permeable. One possible difference is that the objects of study in soft sciences are often complex systems in which variables are not easily isolated and manipulated. As for personality assessment being junk, some undoubtedly is. I should know, since this is my area. We can, however, look at imperfect predictability and generalization as a glass half full rather than empty.
“These diagrams are based not on experiments, but on something totally different: painstaking observation.”
Except that all observations are themselves theory-laden. And the “painstaking observation” to produce a “tidy diagram” of the “circulatory system” took many theory-laden observations, tests, and corrections since the Greek physician Galen (129-216 CE) before it achieved the result that is the currently-preferred theory. There is also the issue of testing theories by various kinds of criticism: not all testing is empirical. There is no way to avoid conjecture and testing by doing theory-free observation instead. We are bound to practice critical-rationalist epistemology whether we understand it or not.
Bryan are you familiar with the work of David Deutsch, philosopher of science and 'grandfather of quantum computing'. Looks at these questions with great insight, I think you'd like his work.
I don’t understand why economists can’t even agree with simple statements such as
Raising the minimum wage will tend to…
Rent control tends to…
Price controls lead to…
Unintended effects of regulation include…
Key institutional arrangements to foster free markets are…
It seems that the path to fame and Nobel prizes in economics is to take easy and obvious answers like the above and instead twist them into exceptions to the rule (which often are politically aligned with the friends and sponsors of said economists). The net result is an embarrassing lack of consistent and reputable knowledge within the discipline. Economics thus looks more like a sorry spin off of bad political science.
Astronomy is another scientific field in which valuable knowledge has been achieved with non-experimental observation. I believe that the same is true in my own area, psychology, specifically personality assessment and measurement, which relies more on correlational than experimental methods. In 1957, Lee Cronbach wrote an article, "The Two Disciplines of Scientific Psychology", in which he discusses the experimental and correlational methods, in which he encouraged combining the two approaches. His 1975 follow-up, "Beyond the Two Disciplines of Scientific Psychology," was pessimistic in tone, concluding that higher-order interactions among variables makes in unlikely that we will be able to make generalizations. Although Cronbach suggested that both the experimental and correlational approaches were equally important, I think he ignores the common perception that experimental science is superior to observational/correlational science (a perception with which I disagree).
Interesting post. It sounds like you’re talking about the perceived superiority of hard sciences over soft sciences, which I think is a useless concept because it ignores the permeable boundaries between the two. I’ve often heard the notion that personality assessment is “junk” science, and I think that’s because it is TOO generalized and people can always find exceptions to the theory. How do you overcome this conundrum?
I agree that the boundaries between hard and soft sciences are permeable. One possible difference is that the objects of study in soft sciences are often complex systems in which variables are not easily isolated and manipulated. As for personality assessment being junk, some undoubtedly is. I should know, since this is my area. We can, however, look at imperfect predictability and generalization as a glass half full rather than empty.
Reminds me of the lack of double blinded randomised control trials investigating the efficacy of parachutes for preventing trauma.
The definition I use for science with my students is this; science is the search for true positive statements.
Observing in economics is a lot more subjective than observing the physical world.
FYI you linked the staff, I assume you meant to link the faculty.
Subjective, yes. But perhaps less SUBTLE. More present/available to the senses.
Science is not experiment. Science is correct prediction. All knowledge is the ability to predict.
“These diagrams are based not on experiments, but on something totally different: painstaking observation.”
Except that all observations are themselves theory-laden. And the “painstaking observation” to produce a “tidy diagram” of the “circulatory system” took many theory-laden observations, tests, and corrections since the Greek physician Galen (129-216 CE) before it achieved the result that is the currently-preferred theory. There is also the issue of testing theories by various kinds of criticism: not all testing is empirical. There is no way to avoid conjecture and testing by doing theory-free observation instead. We are bound to practice critical-rationalist epistemology whether we understand it or not.
https://jclester.substack.com/p/critical-rationalism
Bryan are you familiar with the work of David Deutsch, philosopher of science and 'grandfather of quantum computing'. Looks at these questions with great insight, I think you'd like his work.
I don’t understand why economists can’t even agree with simple statements such as
Raising the minimum wage will tend to…
Rent control tends to…
Price controls lead to…
Unintended effects of regulation include…
Key institutional arrangements to foster free markets are…
It seems that the path to fame and Nobel prizes in economics is to take easy and obvious answers like the above and instead twist them into exceptions to the rule (which often are politically aligned with the friends and sponsors of said economists). The net result is an embarrassing lack of consistent and reputable knowledge within the discipline. Economics thus looks more like a sorry spin off of bad political science.