12 thoughts on ““No Evidence””

  1. On an Open Thread on Instapundit, someone posted a Zoom lecture recording of Czech astrophysicist Pavel Kroupa, who is a contrarian on the whole Dark Matter/Dark Energy hypotheses. What was noteworthy is that Mordecai Milgrom, the MOND dude, was in on the Zoom session and weighs in on some points.

    I think for whoever posted the link, this was not because anybody in the Blogosphere cares whether Dark Matter is for real. Rather, this lecture was a proxy for the science/politics nexus of Climate Change.

    Dr. Kroupa asserted that there is no place for belief in science, or at least scientific investigation carried out by physicists. What there is are assertions or hypotheses or theories, the only thing you can do with them is disprove them by offering evidence to the contrary. And if the evidence is an experimental measurement, the standard is to offer 5-sigma probability of such a disproof.

  2. David Deutsch had an interesting philosophical argument that I’m at least halfway approving of: It’s in the context of his battles with the positivists in physics: (I’m paraphrasing/interpreting here)

    Before you can interpret your sensory data, or any reading from any instrument at all, you have to have a set of theories you are working from about what any of it indicates. You can’t bootstrap yourself from zero beliefs about how the world works to any sort of knowledge about the world – you need to start with some set of explanatory models and occasionally jump to a different set if it better explains the data. Explanation isn’t the same thing as closeness of fit: Ptolemaic epicycles can fit any curve you want, but they don’t explain anything. An explanatory model is generative, has a range to what it can predict from what it was designed to model, and predicts things beyond the original domain.

    One of our difficulties with AI is that we can get them to eventually conform to all sorts of data by beating on a neural network with millions of examples: But what they’ve learned always has a character more akin to a curve-fit than a generative explanatory model. The neural net ends up having expectations, but no explanation: It isn’t doing what we do with our longer ranged understanding of the world. Humans start with all sorts of inborn expectations about how our universe behaves: Object permanence, etc. (ref psychology experiments with babies…)

    1. Explanation isn’t the same thing as closeness of fit: Ptolemaic epicycles can fit any curve you want, but they don’t explain anything.

      I disagree. Epicycles were measurements of the deviations from an existing model (basically an Earth-centric model of the Solar System). There’s not much point to having a model, if you don’t measure how it deviates from the reality it tries to describe.

      That’s what dark matter/dark energy and the MOND theories are about. General Relativity is our current best theory on the cosmological scale, but it fails. These are attempts to model that failure.

      Both for epicycles and for the above cosmology version, you can measure and understand how reality deviates from theory. That is a necessary step for coming up with a better model.

      1. Actually, to be more accurate, epicycles were models of the deviation between the base model and reality. The other theories are of the same sort.

        1. Aren’t epicycles, in effect, a series expansion of an elliptical orbit?

          What is claimed about the Antikythera Mechanism based on analysis and reconstruction is that it predicted planetary conjunctions and eclipses using a pin-in-slot approximate elliptical motion generator, which is similar to epicycles as an approximation?

          1. If you want some fun with epicycles, try reading Maimonides’ Mishnah Toreh, where he tried to get the lunar numbers to work. Oy vey!

      2. I guess I already told my Dark Matter story, but you are going to hear it again.

        I was visiting Fermilab as a tourist in the company of out-of-town family members. Apparently the Tevatron or whatever they called the big accelerator ring is now ancient history because all of the frontier-of-science studies require the more powerful device run by CERN. Fermilab, like many other National Labs, needs “something” to stay in business, as it were, and there were signs in that famous Central Administration Building shaped like a symbol for a sub-atomic particle that they are searching for Dark Matter, using some or other kind of Waiting for Godot experiment of starring at cosmic rays or something and hoping some new particle shows up.

        Our party rode the elevator up to the mid-level Observation Deck open to visitors to get a better view of the large circles in the ground, and in my best rubber-neck tourist at a National Laboratory voice, I announced to other members of my party, “I don’t believe in that Dark Matter thing. I am more of a “MOND person”, myself.”

        A guy who must have worked there was riding the elevator with us, and I think he was busting a kidney to suppress laughing at my remarks, but maybe the last laugh will be on him with Kroupa’s >5-sigma evidence from star cluster and galaxy-group dynamics?

  3. Ah, Santa Ynez! I spent loads of time up there in the 1980s during the Peacekeeper flight test program at VAFB, and got to know a lot of the wineries quite well – though many have gone since then more have emerged. Those were the zinfandel days.

    We even frequented the Hitching Post in Casmalia, though back then it was more like a decrepit saloon that served pretty good (not great) steaks, local wines, and very little else that was good. Nothing like how it appeared in Sideways. I just looked up their menu, and can believe that it is now much more upscale – our crew of young TRW engineers and junior Air Force officers could never have afforded it at the current price range.

    Foxen wasn’t around back then, but they do look like they do interesting things.

    Thanks for evoking memories of “the good old days”!

    1. Peacekeeper, eh?

      I heard of bars excluding patrons with guns.

      Were there any establishments excluding engineers working on MIRV-capable multi-stage solid-fuel rocket boosters?

  4. “No evidence that snake oil works” bothers me. As I understand it, yer travellin’ medicine show featured alcoholic elixirs stuffed with a lot of cocaine and opium, along with assorted mystic herbs and spices and other proprietary ingredients.

    I haven’t done any studies on this, but somebody has. I’m willing to bet that swigging down a bottle of cocaine-laced laudanum will take away any nagging pain and give you a bunch of energy. Probably bring you back the next day for another bottle, too.

  5. Try this on for size everyone. If there are an infinite number of parallel universes the so called multiverse theory it seems to me there is a logical inference from that. Given an infinite number of parallel universes there must be at least one where the random variation of boundary conditions and physical laws roughly equals the JudeaChristian heaven with God in said heaven with God Angels Jesus et cetera as well as hell Sheol happy hunting grounds Valhalla Nirvana and all the rest must logically exist somewhere in said multi verse.

Comments are closed.