Let me toss in a few thoughts on a debate that has been beaten to death, most famously during the science wars in the mid-1990’s. Do scientific judgments depend solely on data? Or do external considerations enter the decision? There are clearly instances (the history of eugenics and fraud in drug trials come to mind) when biases affect scientists’ work. But we acknowledge those case as deviations from an ideal. The question is whether science must involve values.
I’ve found that scientists usually get quite emotional and heated during this discussion. The mere suggestion that our work is not absolutely, perfectly objective really gets our blood boiling. The response is partially understandable. After all much of our credibility depends on the fact that we produce, more or less, objective information. The public counts on the fact that my beliefs about conservation don’t affect my research on climate change. So I do understand some of our indignation.
Nevertheless, I’ve found that we’re often blinded to some basic facts. Consider first the immensity of science. From 1992 – 2002, almost 3 millions papers were published in the U.S. alone. Federal R&D in 2009 stood at almost $145 billion. Corporations added another $290 billion. As scientists routinely brag, the results of modern science have permeated every facet of our existence. It is impossible to avoid.
So when we ask whether “science” is value-free, we’re really asking whether those millions upon millions upon millions of research problems are value-free. Given both how vast and how diverse science is, it’s inevitable that at least some of science is not value-free. Surely some questions are so caught up in the fabric of society that it’s impossible to completely separate science from values. There’s nothing scary about it. Asking an abstract question about “science” goes down down the wrong path. An enterprise as vast as science cannot be easily generalized.
None of this negates the idea that much of science is value-free. We can safely say that quantum mechanics and molecular biology fall in this category. But a lot of regulatory science (e.g. determining the effect of a pesticide) has to be performed under time constraints and limited data. In these cases it’s generally accepted that subjective value judgments play a big role. Heck, the National Academy of Science even wrote a report about it in 1985. Before that, physicist Alvin Weinberg’s “Science and Trans-Science” made the same argument. Sheila Jasanoff’s The Fifth Branch also addresses this topic. There’s actually a fairly large body of empirical evidence proving quite conclusively that regulatory science involves value judgments.
A couple things come to mind as I look back on these debates. First, most scientists (myself included) ironically ignored evidence. All those studies and data on the subject were irrelevant to us. Whenever the subject came up, one of us could have suggested we look at the relevant research. But no one did. Which brings me to my second point.
There’s something about scientists’ training that makes us believe we’re all qualified to speak for “science.” I would never consider speaking authoritatively on condensed matter physics even though I’ve taken a few classes in the subject. Yet in the past I have waxed eloquently about “science.” And the funny thing is that people (especially non-scientists) take me seriously. But if I’m not qualified to speak about all areas of physics, how on earth am I qualified to speak for science?