Tag: scientific models

How Quantum Models Work

Posted on by David Kordahl

by David Kordahl

A notable theorist visits a notable laboratory (Stephen Hawking at CERN, 2013)

The science lab and the theory suite

If you spend any time doing science, you might notice that some things change when you close the door to the lab and walk into the theory suite.

In the laboratory, surprising things happen, no doubt about it. Depending on the type of lab you’re working in, you might see liquid nitrogen boiling out from a container, solutions changing color only near their surfaces, or microorganisms unexpectedly mutating. But once roughly the same thing happens a few times in a row, the conventional scientific attitude is to suppose that you can make sense of these observations. Sure, you can still expect a few outliers that don’t follow the usual trends, but there’s nothing in the laboratory that forces one to take any strong metaphysical positions. The surprises, instead, are of the sort that might lead someone to ask, Can I see that again? What conditions would allow this surprise to reoccur?

Of course, the ideas discussed back in the theory suite are, in some indirect way, just codified responses to old observational surprises. But scientists—at least, young scientists—rarely think in such pragmatic terms. Most young scientists are cradle realists, and start out with the impression that there is quite a cozy relationship between the entities they invoke in the theory suite and the observations they make back in the lab. This can be quite confusing, since connecting theory to observation is rarely so straightforward as simply calculating from first principles.

The types of experiments I’ve had been able to observe most closely involve electron microscopes. For many cases where electron microscopes are involved, workers will use quantum models to describe the observations. I’ve written about quantum models a few times before, but I haven’t discussed much about how quantum physics models differ from their classical physics counterparts. Last summer, I worked out a simple, concrete example in detail, and this column will discuss the upshot of that, leaving out the details. If you’ve ever wondered, how exactly do quantum models work?—or even if you haven’t wondered, but are wondering now that I mention it—well, read on. Read more »

The Limits of Conspiracy Debunking—Revisited

Posted on by David Kordahl

by David Kordahl

The sides of a Roman die (image from Wikipedia).

Note: This piece is an accidental addendum to my column of March 2021, “The Limits of Conspiracy Debunking,” though it can be read separately.

Sometimes, we’re surprised. Though everyday surprises can be comedic, the surprises that we register collectively are more often tragic. My parents both remember the assassination of John F. Kennedy as one of the most shocking events of their childhoods. I suppose the attacks on the World Trade Center and the Pentagon on September 11, 2001, constitute the most shocking of mine. Both the JFK assassination and the 9/11 attacks have attracted conspiracy speculation ever since they occurred. And there are good reasons for this, I contend, even if no conspiracies were involved.

Collective feelings of surprise, of widespread shock, reflect a vague feeling that such events were unlikely, though their very unlikeliness makes their odds hard to calculate. In such cases, alternative explanations—“conspiracy theories,” if you’re feeling ungenerous—seem attractive because they change our estimated likelihoods for surprising events. After these events, it’s natural enough to ask, Why didn’t I see that coming? We might consider, with some shame, whether our expectations should have included wider possibilities, so we might have been less surprised.

After such shocks, the prophets who forewarned disaster gain legitimacy. People retrospectively consider alternatives that accommodate their prior surprise. This rethinking serves to change their subjective odds for the likelihood of the original event. But that’s a problem, if we’re interested in any sort of self-consistency, since this retrospective modulation of odds is only reasonable if we actually should not have been surprised.

This may all sound circular, but there’s a solution at hand. If unique events can be reclassified as non-unique, we can move away from seeing events as being unprecedented, and statistics will once again apply. Returning to shocking events, JFK might be rolled into the category of “political assassinations,” and 9/11 into the category of “political terrorism,” and we might then think through the odds by examining trends in those categories. Read more »