Electrodynamics: Comparing Purcell & Griffiths Textbooks

[Jianphys17]

Hi, I've 2 em texts: Purcell, ( Electricity&magnetism ) and the Griffiths (intro to electrod.)
On which of the two, is best to study ?

 

[Vanadium 50]

Both.

 

[pixel]

It's kind of a personal decision as to which text matches your style better. Pick one, and if there's a topic that doesn't seem clear, read about it in the other one.

So I'm agreeing with "both."

 

[Jianphys17]

Thank you for help !

 

[dextercioby]

I think there are two types of books: the one by Purcell is old (written from the Univ. Berkeley general physics course at the beginning of the 1960s, very many words compared to formulas, no tensor formulation in Minkowski spacetime, not all explanations are clear), while the one by DJ Griffiths has everything into one book without the fancy mathematics by JD Jackson. There's no reason for me to choose the book by Purcell over the one by Griffiths.

 

[vanhees71]

If you have only the choice between these two, I'd definitely recommend Griffiths...

 

[Demystifier]

I have a special emotional attachment to Purcell's Electrodynamics because I studied this book before going to college during my serve for the former Yugoslav Army, just a year before the war in Yugoslavia started. Purcell helped me to not get crazy in army. So don't tell me anything bad about Purcell.

 

[ibkev]

the one by Purcell is old

Do you have thoughts on the big update to Purcell that was done by Morin? He seems to have made a lot of changes: http://www.people.fas.harvard.edu/~djmorin/PMpreface.htm

 

[dextercioby]

It is suitable as a companion to a general physics textbook (Serwey, Tipler, Eisberg&Resnick). Yes, I know that American schools like to make a distinction between undergraduate and graduate, nonetheless, I still prefer a more modern text. Actually, to be really fair, the first exposure to components-based tensor calculus is made during the electrodynamics course (at least in my country, in which the modern official university textbook is based upon JD Jackson's EM text in his 2nd - 1975 edition).

If you rewrite a textbook, polish some phrases, make them clear, then at least bring it to physics newer than 1915. As you know, the full blown tensor formulation of electromagnetism was available to David Hilbert as late as November 1915 and his (in)famous article on variational principles for the gravitational field.

 

[Buffu]

Resnick

My high school textbook was more in depth than Resnick's book. I don't get for whom this book is for ? middle schoolers or kindergarteners ?

 

[ibkev]

My high school textbook was more in depth than Resnick's book. I don't get for whom this book is for ? middle schoolers or kindergarteners ?

It's uncommon for high school textbooks to expect the reader to know calculus as Resnick does.

 

[ibkev]

Yes, I know that American schools like to make a distinction between undergraduate and graduate, nonetheless, I still prefer a more modern text.

I would say that Canadian schools make a similar distinction - so I'm curious now - do other parts of the world have a different approach? I can't imagine Jackson or say, Goldstein, being given to 2nd year undergrad students but perhaps I need a better imagination? :)

 

[vanhees71]

It is suitable as a companion to a general physics textbook (Serwey, Tipler, Eisberg&Resnick). Yes, I know that American schools like to make a distinction between undergraduate and graduate, nonetheless, I still prefer a more modern text. Actually, to be really fair, the first exposure to components-based tensor calculus is made during the electrodynamics course (at least in my country, in which the modern official university textbook is based upon JD Jackson's EM text in his 2nd - 1975 edition).

If you rewrite a textbook, polish some phrases, make them clear, then at least bring it to physics newer than 1915. As you know, the full blown tensor formulation of electromagnetism was available to David Hilbert as late as November 1915 and his (in)famous article on variational principles for the gravitational field.

Well, classical electrodynamics has been completely formulated in the years around 1908, when Minkowski gave the relativistically covariant formulation. I'd say that's the most modern way you can and should present the subject (augmented with some general principles of gauge theories). I also find it pretty unsatisfying that also nowadays still new textbooks appear that are written in the style of the 19th century (with the only exception that of course the vector notation is used today). Jackson is an example for this, but it's still among the best and most comprehensive books on the subject. Of course the 2nd edition is a bit clearer because it uses the Gaussian system of units throughout, which is the right choice (although I'd prefer Heaviside-Lorentz units, i.e., rationalized Gauss units) for theoretical electromagnetism the 3rd (in Germany although the 4th) edition uses the SI in the non-relativistic part of the book and Gaussian units in the relativistic one, which is very telling.

The sad thing with Purcell is that it is in principle way more modern in principle, but that it fails in giving a clear exposition of the relativistic point of view in its attempt to be pedagogical and avoiding to introduce the machinery of tensor analysis in Minkowski space first and using instead some confusing ideas to derive electromagnetism from electrostatics plus some handwaving introduction of relativity.

The most modern and straight-forward text with regard to the relativistic approach from the beginning is still Landau and Lifshitz vol. II (also 2-3 decades old). Another good text is vol. 3 of Scheck's textbook, emphasizing a bit more the gauge-theoretical aspect.

What's infamous about Hilbert's article on GR? He's the one who introduced the action principle. That's it or what are you after?

 

[dextercioby]

Hilbert's article on GR is based on Mie's ideas, as far as the interpretation is concerned. By our current mind, we could call Mie a crackpot.

See here: http://www.tau.ac.il/~corry/publications/articles/pdf/mie.pdf

 

[vanhees71]

Interesting. I'd not call Mie a crackpot per se. At the time many physicists and mathematicians were looking for a unified worldview based on electromagnetics (and gravity). Another example is Weyl's theory introducting electromagnetism by gauging the conformal invariance of GR without matter. It's of course wrong, as has been told him by Einstein and Pauli immediately after he presented the idea to them, but he didn't listen to them. It's sometimes really amazing, how brilliant mathematicians can so obviously fail to recognize the consequences of their ideas. Ironically till today we call the idea "gauge theory", of course not "gauging" conformal symmetry but global symmetries like phase invariance (electrodynamics as U(1) gauge theory) or non-Abelian extensions (Yang+Mills building the mathematical foundation of the Standard Model)) anymore. So also wrong ideas can become important methodological concepts!