Riordan replies: I appreciate Robert Reiland’s critique of my Opinion piece and the opportunity it gives me to elaborate. I agree that what we call a general theory, such as quantum chromodynamics (QCD), can never be proved absolutely true in all possible cases. But after a long string of successful tests, a theory is generally considered true, and it will have to be incorporated in any theory that may supersede it. For example, the somewhat more limited statement that quarks exist will always be true in some context. It has become a “scientific fact.”
Yes, the quarks we recognize today are substantially more than what Murray Gell-Mann predicted in 1964. Our physical idea of quarks has been extensively elaborated since then—especially by the emergence of QCD, as Reiland notes. Theories indeed play the central role in establishing the physical meaning of our concepts, and thus their meaning evolves. But I do not agree that they play the only role in this normal historical process.
Even when today’s dominant standard model of particle physics is replaced by a wider, better theory in the near future, as many physicists expect, I am certain that quarks will continue to exist in some sense. Protons and neutrons did not cease to exist just because we discovered quarks in the 1960s and 1970s. But our physical picture of them changed dramatically.
Despite Juan Pablo Pardo-Guerra’s objections, I will continue to assert that theorists who generate physical models that cannot be reasonably tested are engaging in metaphysics rather than physics. But I am not, as he claims, denigrating the philosophical underpinnings of science, for which I have a high regard. Where he uses “metaphysics,” I prefer the word “philosophy,” which for me has a slightly different, broader meaning.
As Gerald Holton and other historians of science have noted, our deep philosophical predispositions have long guided theory choice—even the selection of what physical problems are worth addressing. My own favorite scientific philosophy happens to be pragmatism. Although it may not be an a priori better choice than other philosophies, it is certainly a more effective philosophy in that worthy scientific theories are required to have observable consequences. They have to do things, not just be. Most practicing scientists share this philosophical prejudice, consciously or not.
In my Opinion piece, I protested what I see as the emerging divorce of theory from experiment: Mathematically adept theorists are increasingly publishing ideas that may have no observable consequences.
And like any divorce, this one can originate on either side. Raw Baconian empiricism, devoid of theoretical interpretation, is also not science. Perhaps I was a bit too strident in championing experiments in my article; I seem to have been misinterpreted as being against theory. Modern physics, I admit, cannot exist without it.
What natural philosophers developed in the 17th century, the marriage of ideas and observations, is an immensely powerful intellectual process that has radically altered human activity and, with it, our physical landscape. It has sometimes been a difficult marriage, but it has been a very rewarding one, too—and one that physicists must continue to live with, for better or for worse.
