Popper upside down

by Neil Rickert

The idea for the title comes from a blog post by Sabine Hossenfelder:

Hossenfelder is herself a physicist.  She makes a good argument on the problems with the way that scientists use (or misuse) Popper’s idea of falsification.  It is well worth reading.

For myself, I slightly disagree.  I don’t think she should blame the problem on Popper alone.  The problem is bad philosophy of science.  And, unfortunately, there is a lot of bad philosophy of science.  Some of that bad philosophy comes from professional philosophers.  And some of it comes from scientists themselves.

[Yes, this was a very short post intended to reference Hossenfelder’s post.]

Advertisements
Tags:

5 Comments to “Popper upside down”

  1. I simply couldn’t agree that predicting a certain person would kill the Donald by a certain date is a testable prediction.

    A testable prediction would be that after x and y happen, person z will make the decision to assassinate and start making plans. That is testable because you can follow events x and y and then monitor person z for the behaviour change.

    Predicting event y will happen on or before date w is not a testable prediction because all you can do is wait for the date to pass, you can’t test for it.

    After that I struggled to read further.

    Like

    • I simply couldn’t agree that predicting a certain person would kill the Donald by a certain date is a testable prediction.

      You wait until that date, and see if it happened. That tests the prediction.

      You prefer prediction of the form: if I do action X, then event Y will follow.

      Let’s call the first of those (the Trump example) the bad kind of prediction. And let’s call the second to be the good kind of prediction.

      Hossenfelder is saying that too much of modern particle physics is based on the bad kind of prediction.

      I’ve been troubled by particle physics for some time, though I’m not into physics deeply enough to be able to make that kind of criticism.

      My way of looking at it, is that science should be about measurement and observation. But once you get below the threshold of the Heisenberg Uncertainty principle, what is being observed and measured? At that level, observations are noise. There can be statistical trends in the noise, and those can be studied. But it did seem to me that physicists were analyzing the noise to see if they could find indicators of “the god particle” in that noise. And it isn’t at all clear to me that they are doing anything useful.

      Hossenfelder is describing the same sort of issue in terms of prediction. She thinks that what I called “the bad kind of prediction” is more like prophecy than prediction.

      Like

      • “Hossenfelder is saying that too much of modern particle physics is based on the bad kind of prediction.”

        I clearly didn’t read far enough. Assuming she is correct, I would agree that that is a bad thing and should be called out. Did she give a specific example to demonstrate?

        I’m still hung up on her claiming that the ‘bad sort’ is testable. If it is testable, then what’s her problem with it? Testable is good right? This is why I gave up reading, I thought her opening section was a little contradictory.

        I think it’s not testable because the event has a single opportunity, it can’t be repeated under different circumstances. That doesn’t meet the criteria I consider when something is called testable. Maybe that’s the issue, she and I have different definitions of what it means to be testable.

        “My way of looking at it, is that science should be about measurement and observation.”

        Yep!

        “But it did seem to me that physicists were analyzing the noise to see if they could find indicators of “the god particle” in that noise. And it isn’t at all clear to me that they are doing anything useful.”

        Fair comment and my knowledge isn’t good enough to determine if the examination method was good or bad.

        Like

        • I’m still hung up on her claiming that the ‘bad sort’ is testable.

          It’s not so much her claim. Yes, it is testable, but only in a weak way (as you have noticed). Hossenfelder’s point is that some physicists are using that kind of weak testability to say that their work is falsifiable and is thus good science. And she is objecting to that (and rightly so).

          Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: