I am often struck by how historical accounts of science say so little about experiment. Think of Arthur Eddington’s famous measurement, for example. Paragraphs about Einstein and friends weave a rich narrative about how general relativity predicted something that had never before been expected: that starlight should bend around the sun. Such tension must surely culminate in some incredible triumph — but no; “This was observed by Eddington in 1919.”
It is as if a child could have done it. A child without a telescope. A child who did not have to haul apparatus all the way to a jungle- covered island off the western coast of central Africa, and assemble it with hardly any infrastructure, and without craftsmen and precision workshops to repair any damage or make major adjustments. It’s as if this task required no more than a child without special skills, certainly not the skills that go with being Astronomer Royal.
I guess that Eddington must have just gone along for sightseeing while some unnamed and forgotten urchin did it. Obviously not what the author meant, but fully accommodated within what the author wrote.
There is a similar story in physical optics. Jean-Augustin Fresnel extends Thomas Young’s wave theory of light to elegantly cover a wide range of optical phenomena. Then Siméon-Denis Poisson takes it a step further, cleverly deducing a reduction ad absurdum: if Fresnel’s theory is right, then in the middle of the shadow should be a pinprick of bright light. The narrative escalates into a moment of unresolved conflict, and then, like a ruined soufflé, collapses: “observed by Arago in the same year.”
You’d think that it could have been seen by anyone. No special apparatus. No special skill. Maybe it really could have been seen by anyone: “It had been previously observed by Maraldi, who did not realize its significance.”
So not only was Maraldi as skill-less as Arago, but he was also stupid. Somehow I doubt that this is what any author means but, as in Eddington’s case, the deliberate lack of detail can easily be read as damning by faint praise.
Why is empirical work so often trivialized as if it requires no special knowledge, and as if it serves merely as a formality, confirming what everyone already knows?
Could it be that some of us do not actually believe empiricism to be a hallmark of modern science? Or that it is a hallmark in an even more literal sense of being stamped on after the fact following a quick, routine evaluation? I doubt that very much. But, regardless of what we claim to believe, it is what histories of science — whether by historians or by scientists — very often portray.
Unfortunately, I see that dismissal of empiricism also in how science is taught. Lectures. Problem sets. All pen and paper, with the labs serving only as a supporting act. If, that is, they are required: in some institutions, lab courses are optional. In some schools they do not exist, and no one sees fit to fund them. Where they do exist, they seldom teach experimental skills and cognitive habits, but instead relegate experiment to ancilla philosophiae.
Even if you do not care about the nature of science, this matters. Only in experiment do you learn what skills and knowledge are required to get empirical evidence, and how even the finest skills and knowledge are curtailed, constrained and redirected by circumstance. Measurements and observations are biased. Accuracy and precision are limited. Experimental results cannot be fully interpreted without knowing all that. This does not apply only to scientific data, but to observation and measurement in life at large: what do you make of the percentage of students who pass or fail? The reported results of medication trials? The fuel consumption efficiencies of two cars that you are thinking about buying?
It seems to me that the implicit trivialization of experimental work does a serious disservice not only to science, but to liberal education overall.
