Wednesday, March 27, 2013

Strassler On Talking About Science

I think it very important for scientific experts to be clear, when they speak in public, about what is known and well-established, what is plausible and widely believed but still needs experimental checks, and what is largely speculative and could very well be false. (For example: The Higgs particle and field are nearly established; inflation is increasingly plausible; any connection between them is speculative.) . . .
Just as we widely agree the Higgs particle must have zero spin, and that the inflaton is quite likely to have zero spin, I’d like to see a consensus emerge that public communication of particle physics, string theory and cosmology should also have zero spin. Too bad that’s still a rather speculative idea.
From Matt Strassler's blog.

His point is well taken.  I'd would draw the lines between some of the categories he identifies in moderately different places than he does, but his framework is a sound one. 

I would suggest, however, that in many cases there are two or three competing positions which are plausible and widely believed by a subset of the scientific community which are not "largely speculative", and that in those cases (e.g. SUSY, MOND and loop quantum gravity), it would be helpful to acknowledge that there are competing theories and to explain their relative levels of acceptance. 

Often there will be a majority or plurality position, and one or more minority views held by significant numbers of respectable mainstream scientists, which have not been resolved and may be impossible to resolve for extended periods of time due to a limited experimental data.  In these cases, all off the competing theories generally produce very similar predicted phenomenological outcomes within the range of experimental data whose accuracy is not seriously subject to question.  indeed, sometimes these differing positions arise from disputes over the validity of alternative experimental methods.

Likewise, it is often important to distinguish between largely speculative ideas that are professionally respectable ideas, even if they are not widely believed by any subset of the scientific community at this point, and "crackpot" ideas that are starkly contradicted by widely accepted empirical evidence and are contrary to widely accepted physical laws, or are internally flawed in deep ways (e.g. they are not mathematically or dimensionally consistent).

This is particularly important in the area of fundamental physics, where the proportion of all published work of professional physicists that is largely speculative is much larger than in many other academic disciplines. 

A huge share of the published work in theoretical physics analyze toy models of possible laws of nature that are largely speculative or even known to be contrary to empirical evidence, but are not "crackpot" ideas.  They are published with an eye towards understanding the implications of that class of mathematical models to see if they could possibly be made to correspond to empirical evidence if further developed, and to determine what implictations the "new physics" in those models beyond the Standard Model and general relativity might show.  These papers are intermediate steps in the massive undertaking of looking for a final and complete set of the laws of nature and are driven by the known imperfections, at least from a point of view of a comprehensive and rigorous set of laws of nature, with the status quo.  But, for the most part, they don't even pretend to be plausible and widely believed evidence based inferrences about the way that the world already is right now.

A person not familiar with this state of the published and peer reviewed literature in theoretical physics could easily be led astray into thinking that largely speculative ideas have more importance than they actually do, because these kinds of published and peer reviewed papers setting forth largely speculative ideas are far more rare in many other academic disciplines.

I'd also note that it is often the case that we know with a great deal of confidence that some scientific proposition is X or Y, but don't know which one is correct.  The range of possibilities may be "known and well established", but statements about which of the possibilities is actually right may be "largely speculative."

No comments: