I should really know better than to click any tweeted link with a huff.to shortened URL, but for some reason, I actually followed one to an article with the limited-reach clickbait title Curious About Quantum Physics? Read These 10 Articles!. Which is only part one, because Huffington Post, so it's actually five articles.
Three of the five articles are Einstein papers from 1905, which is sort of the equivalent of making a Ten Essential Rock Albums list that includes Revolver, Abbey Road, and the White Album. One of the goals of a well-done list of "essential" whatever is to give a sense of the breadth of a subject, not just focus on a single example, so this is a big failure right off the bat.
But it's even worse than that, because none of the three 1905 articles is the photoelectric effect paper, which is the only one of the lot that has any quantum physics in it. There's a fourth Einstein paper on the list, as well, the theory of general relativity, which is famous for not being compatible with quantum mechanics. So this is really like a list of Ten Essential Rock Albums that includes three country songs and a Bach concerto.
I thought about using this as an opportunity to generate a better Ten Essential Quantum Papers list, including stuff like Bell's Theorem (the physics equivalent of the first Velvet Underground record) and the No-Cloning Theorem (the physics equivalent of punk rock) (brief pause to let those who know Bill Wootters try to reconcile that mental image). And if you would like to make suggestions of things that ought to be on such a list in the comments, feel free.
(I'm also open to suggestions of better musical analogies-- maybe the EPR paper is the real Velvet Underground record? With Bell's paper being punk rock, making Wootters and Zurek... Nirvana, maybe? Or maybe Shor's algorithm is the "Smells Like Teen Spirit" of quantum physics (Again, a brief pause while those who know Peter Shor try to picture him as Kurt Cobain)...)
But, really, on reflection, the whole exercise is kind of silly even by the standards of clickbait blog topics, because that's not how science works. In science, and particularly a highly mathematical science like physics, there's not that much real benefit to reading the original source material. The best explanation of a central concept is rarely if ever found in the first paper to present it. This goes right back to the start of the discipline, with Newton's Principia Mathematica, which nobody reads because it's written in really opaque Latin, a move he claimed in a letter was deliberate so as to avoid "being baited by little smatterers in mathematics" (Newton was kind of a dick). Newton's mathematical notation is also pretty awful, and I've heard it claimed that the reason physics advanced faster in mainland Europe than in England during the 1700s was that on the continent, they adopted Leibniz's system, which was way more user-friendly and is the basis for modern calculus notation. Similarly, Maxwell's original presentation of his eponymous equations is really difficult to follow, and it's only after the work of folks like Heaviside that they become the clear, elegant, and bumper-sticker-friendly version we know today.
That's not to say that there's no value in reading old papers-- I've had a lot of fun writing up old MS theses from our department, and older work can be fascinating to read. But unlike primary works of (pop) culture, they're much better if you come to them already knowing what they're about. The fascination comes from seeing how people fumbled their way toward ideas that we now know to be correct. It's rare for a "classic" paper to get all the way to the modern understanding of things, or even most of the way there-- most of the great original works contain what we now know to be errors of interpretation. Others are revered today for discoveries that were somewhat tangential to what the original author thought was the main point-- the Cavendish experiment is thought of today as a measurement of "big G," but he presents it as a determination of the density of the Earth, because that was of more pressing practical interest at the time.
If you want to learn science, you're much better off looking up the best modern treatment than going back to the original papers. A good recent textbook will have the bugs worked out, and present it in something close to the language used by working scientists today. A good popular-audience treatment (ahem) will cover the basic concepts starting from a more complete understanding of the field as it has developed, and with an eye toward making those concepts accessible to a modern reader. It's not foolproof, of course-- the steady progress of science over a stretch of decades often means that newer books need to cover a huge amount of material to get to the sexy cutting-edge stuff, and sometimes scant the basics a bit. But by and large, if you're curious about quantum physics, you'd be much better off hitting the physics section of your local bookstore or library than digging through archived journals for the original papers.
So, a list of "Ten Essential Papers on Quantum Physics" is a deeply flawed concept right from the start, at least if the goal is to learn something about quantum physics that you didn't already know. The same is true of almost every science, with a few exception-- Darwin's On the Origin of Species is still a really good read, but it's the exception, not the rule. Such a list can be useful as a sort of historical map, or for providing some insight into the thought processes of the great scientists of yesteryear, and those can be very rewarding. But if you're curious and want to learn, I don't think any original papers can really be considered "essential."
- Log in to post comments
" It’s rare for a “classic” paper to get all the way to the modern understanding of things, or even most of the way there– most of the great original works contain what we now know to be errors of interpretation."
I think Einstein's papers are a significant exception. Those that I have read are models of clarity and have aged extremely well. Some are quite accessible to an advanced undergraduate. My top 3:
Brownian motion (1905)
The quantum theory of radiation (1917) (at least the 1st part)
The EPR paper (1935) - although he's not the sole author.
Interestingly, Einstein later complained that the EPR paper didn't quite capture the argument he intended. He said Rosen did most of the writing of that, because it was in English.
The 1917 photon paper is a really nice piece of work. Kleppner's Physics Today from 10-ish years ago on that is really nice, too.
This was exactly the mistake made by Mortimer Adler and co. when they assembled the Great Books collection that was sold by the Britannica people for so many years (my parents bought it back in the early 1960s, its heyday). The "Great Books" concept makes sense for fiction, although everyone argues about the details (such as who should be included). But if you want to read Joyce, Austen, Conrad or Hemingway, you go and read them.
But Adler included original works from Newton, Harvey, Lavoisier, Huygens and so on (even Ptolemy). And for exactly the reasons you've written about, this is an insane way to learn what they had to say. It also gives any readers (and I'm not sure how many there really were, in this case) no chance at all to put these works in any sort of context. Even when I was twelve years old and taking some of these books off the shelves, I knew that there was something funny about having four-hundred-year-old science up there. Adler, though, was no scientist (and how), and never seemed to realize the problem.
What do you think about Feynman physics lessons ?
I started to read it (mechanic 1 and 2, the other one are waiting), i'm not physicist but biologist (molecular type) and i found them excellent. I mean he just blow my mind and (i think) i understand most of what he said thank to the fact that i'm not too bad at reasonning and math and because he is very sound and clear.
Do you think those book are still good (up to date) for learning physics ?
We did do a Classic Papers in Planetary Science seminar in grad school, but some of the work of each day's presenters was to not only read the paper but to dig up the context, what's changed since then, and how the information was used. Which made the course interesting, both to learn how things were done and to see how a paper, even an important one, fit into a body of work.
Many of the students also knew something of the field already, so it was more interesting for history and cultural reasons than scientific.
But if you want to read Joyce, Austen, Conrad or Hemingway, you go and read them.
Sure, it works for authors who write in a recognizably modern version of a language in which you are fluent. Try this with Shakespeare, or with Dostoyevsky (unless you happen to be fluent in Russian), and it doesn't work so well. There is a reason most high schools use editions, such as the Folger Library editions, that include marginal notes: many words are used differently today, or have disappeared from English in the last 400 years. (Shakespeare himself invented several thousand words; about half of them are still used.) Translating a work of literature invariably leads to ambiguities and shifts in meaning or association, because exact translation is often not possible (especially with literary language).
The point here is that reading Shakespeare, or Dostoyevsky in translation, is more like reading the original Principia or Maxwell's work. You can get some of the point, but without a guide you tend to lose some important details.
As for musical analogies, I'll mention one of my recently acquired albums. The artist is David Garrett, a violinist whose repertoire includes both popularizations of classical tunes and arrangements of modern popular music. The iTunes music database classifies this album's genre as Classical, which makes sense for some of the songs on it, e.g., the one based on the scherzo movement of Beethoven's Ninth Symphony. The album also includes a piece by noted classical composer (and Ph.D. astrophysicist) B. H. May called "We Will Rock You". Calling this song "classical music" is like calling Einstein's GR paper an essential quantum physics paper.
On the subject of learning from the classics, I've never met anybody who went to college at the great-books St. John's schools who went into science. (The only person I've met who I know went there was, quite literally, barking mad. I don't think he's really the best example, though.) I wonder what that program is like, and if it's even possible to go on to science grad school from there without making up a bunch of stuff elsewhere.
#6, Eric Lund:
You're certainly right about translation (that's why I didn't include any non-English authors in my example list). Shakespeare is definitely improved for a modern reader with extra notes as well.
But these are works of art: how they're composed and phrased is a key part of their impact. Not so with scientific papers - it's all about the content. And over the years, that content is added to, amplified, and refined in a way that just isn't like what happens with an annotated edition of Shakespeare (e.g.) What Einstein said doesn't depend on Einstein saying it - but what Shakespeare said does.
I do know someone that went to St John's and wound up getting a Ph.D. in physics. He did have to make up some undergrad classes before moving on.
Ur doing it wrong. You should be working on something important, like a web site that lets readers ask the question "Which seminal quantum physics paper would I be?" basing the outcome on inane questions about favorite color, which muppet is best to be stuck on a desert island with, etc.
From now on, it will be hard not to think of the 1927 Solvay Congress in Brussels as the QM equivalent to Woodstock.
IMO, the only important part of the Principia for those of us teaching physics is the long discussion of the First Law. We have to deal with Aristotelian notions every day in first-semester physics, and Newton had to find a way to dispense with more than a millenium of actually teaching them. He ends up definining them away by (in more modern terms) requiring the choice of an inertial coordinate system where inertial forces (vis inertiae) were required to keep things going but were NOT part of the analysis and thus not called "forces". It is a clever trick that I have never tried to use when teaching but wonder about trying with the more stubborn misconceptions.
Also, +infinity to the suggestion @ #10, and Solvay 1927 was Woodstock 1968, although I would lean towards Monterey 1967.