Tobias Schottdorf    About    Archive    Feed

Sabbatical Reading List: The Feynman Lectures

I’m on sabbatical, so I get to look into things without the distractions of my day-to-day job. This is one of them.

Feynman was not a saint1, but without doubt one of the geniuses of the 20th century. And I am, without doubt, someone who in their confused highschool years took the genius decision to opt out of physics once that was possible, presumably to spend more time staring at a computer screen).

Three months ago, with half a year of sabbatical at my disposal, I found myself picking up (i.e. navigating my browser to) the freely available Feynman Lectures on Physics2 and have since had trouble putting them down.

Feynman was sometimes referred to as “The Great Explainer”, referring to his uncanny ability to decompose and boil down the complicated until it looked simple and digestible, for leaving no stone unturned, and for playfully handholding his audience to a deep understanding of the matter at hand, one toy example at a time.

Hopping on Feynman’s bandwagon works well only to some extent, though. What you get are lucid explanations and great intuition about physical processes, with lots of little examples to really make you understand how it works.

What you don’t get are solid proofs. You could say, who needs a proof when intuition makes it all clear as day? And I, with a doctorate in a heavily physics-influenced part of pure mathematics3, would agree and trust myself to be able to spot sketchy arguments.

But - in volume two of the same Feynman Lectures, Feynman got the Faraday cage4 shielding effect completely wrong: He predicts shielding exponential in the mesh gap, while it’s in fact linear5. The mistake is assuming that the wires (discs in 2D) of the cage can be made infinitely thin. But going from constant-voltage discs to point charges changes the question you’re answering.

That is one thing, but we all make mistakes. The other caveat is that once you get into the quantum world, you will have to face the fact that Feynman was fundamentally a “particles guy”, which is problematic because particles are just a sometimes convenient way of talking about fields which are highly localized, but not a fundamental concept.

Assuming the particle perspective when that doesn’t apply brings up apparently paradox situations such as that observed in the double-slit experiment, which isn’t a paradox when you realize you’re observing fields. In short, there are no particles, there are only fields6, and Feynman appears to be on the wrong side of teaching history (while certainly having been on the right one scientifically). At least, as the article6 points out, most of teaching is on the wrong side of history on that point, presumably due to the fact that particles come naturally to humans and even more so when you enter the quantum world through classical Newtonian mechanics.

There also aren’t any exercises in the original lectures, though a list is maintained at CalTech, along with errata and notes2. For more complete coverage, a book exists7, but reviews point out that many problems are presented without a solution, which can be off-putting.

With that in mind, I have so far passed many hours reading my way towards the end of volume one of the lectures and, so far, I can strongly recommend them. Imagining myself in my first semester in this class (as it was originally taught) would’ve surely steamrolled me.

But maybe that’s what it means to study physics?

  1. The problem of Richard Feynman

  2. feynmanlectures.info  2

  3. Partial Differential Equations, more precisely the analysis of nonlinear dispersive equations, and even more precisely, a result on some quadratic Klein-Gordon equations

  4. Faraday cage (Wikipedia) 

  5. Surprises of the Faraday Cage, Lloyd N. Trefethen, 2016. 

  6. There are no particles, there are only fields, Art Hobson, 2012.  2

  7. Exercises for the Feynman Lectures on Physics on Amazon.com.