One of the things that bother me the most about book* recommendation requests is the endless stream of similar requests that can spout around any specific topic, for all the different possible reader levels and prerequisites.

"I'm an undergrad and I want to learn QM" says someone. "Me too, but can you keep it to minimal functional analysis?" says someone else. "Actually, can you cut down on the linear algebra?" and "I haven't taken analytical mechanics, so could you go easy on the hamiltonians and lagrangians?" say two more people...

...and on and on it goes. It is completely impractical to have a new thread for each such request: we should have a single, canonical book-recommendation thread for, say, ~undergraduate quantum mechanics, and close all the requests in the paragraph above as duplicates of the main thread. I see this as absolutely necessary to avoid the quality dilution and the clutter that would come otherwise, and which led us to ban the tag to begin with.

However, it is unfair on people coming to this site for book recommendations to give them no voice regarding what level of books are present in the answers to the canonical thread. If I do find the main thread and find it to be full of books that need no functional analysis, but I can't handle the linear algebra in them, I should still be able to invite answers at that level.

I understand I'm a little late to the game, but I was busy working on other policy for the site. (And, you know, my PhD.) I nevertheless would like to propose a change in how the recently adopted policy works on the question side of the affair. (The answer side is, on the other hand, well dealt with, I think.)

I propose we keep a single thread per topic, and that the specific requests of level be included as edits and comments to the answer.

Thus, in this scheme all the requests above would be merged into a single thread which would be formatted along the following lines

Good undergraduate quantum mechanics books

What are good books or online resources to learn quantum mechanics at an undergraduate level?

Level of resources

If you are looking for resources in this topic but you find the answers in this page too advanced or too basic, add a request in this list.

  • Are there good books that use minimal functional analysis?
  • I'd like resources with very little linear algebra.
  • What are books with little resource to lagrangians and hamiltonians? My analytical mechanics is a little shaky.

This proposal does require additional curatorial attention, in deleting parts of the resource-level list as they get addressed and newer requests are added. However, I feel that this is quite small in comparison with the bulk of work of curating the answers. It provides input to the people who have very specific requests, while keeping the answers condensed into one big resource that's useful for future visitors, instead of scattered into lots of small, specific lists.

Finally, as I have argued elsewhere, policy changes are only half policy changes if they don't come with a documentation proposal. Thus, this change could come with the following paragraph added to the policy.

How do I ask a resource recommendation question?

First of all, search the site to see if the topic has been asked before. Adding the tag [books] to your search may be useful. If the topic has not been covered, start a new question. Keep it general! We want questions and answers to be useful for future visitors. Ask for a specific topic. If you have a specific level or set of prerequisites that you'd like answers to address, add it, separately, at the end of the question.

If the topic has been covered, don't ask a new question! If the resources in that thread are too advanced or too basic for you, or you have some specific set of prerequisites that the current answers don't address, edit the current question to request it. Add your request to the list provided, or start one like the one in this question.

What say ye?

*To appease the purists, throughout this post "book recommendation" should be understood to mean "resource recommendation".

  • $\begingroup$ But looking at your bolding I feel that that was indeed the intention of the policy, the level is mentioned in the answers but the question is at an appropriate level of just-a-bit-broadness. I recall mentioning something similar somewhere, hold on... $\endgroup$ Commented Nov 15, 2013 at 21:22
  • $\begingroup$ Yes. But we need to provide a way for people to request new level-of-resources, and to make it evident to new users that they can ask for such. (As I said, I'm very happy with the answers side of the policy.) $\endgroup$ Commented Nov 15, 2013 at 21:23
  • 1
    $\begingroup$ Ah, you're trying to make it evident to new users. In that case, I'll comment after reading the post fully, when I get the time (and when my mind is awake) :) $\endgroup$ Commented Nov 15, 2013 at 21:25
  • 1
    $\begingroup$ I think the intent of the new policy is that a good answer should cover books at several levels. E.g. "X is good if you're a complete beginner. Y is good if you want to see precise mathematical proofs. Z is good if you want a coordinate-free treatment." etc. Basically, the answer enumerates the standard references so that anyone who wants a book on the topic at any level should find a decent match in the list. $\endgroup$
    – David Z
    Commented Nov 16, 2013 at 6:45
  • 1
    $\begingroup$ @DavidZ Well, not totally, it's OK for a single answer to just mention one book at one level, however the overall corpus of answers should accomplish what you say. $\endgroup$ Commented Nov 16, 2013 at 8:43
  • $\begingroup$ @DavidZ I agree that that should be the goal. However, the current policy does not cater to people who did not find references at their level in the current answers. I also agree with Manish in that there should not be any pressure on anyone to provide more than one book in their answer. $\endgroup$ Commented Nov 16, 2013 at 19:11
  • $\begingroup$ @Manishearth I thought the entire point of the new policy was that that was not OK. $\endgroup$
    – David Z
    Commented Nov 20, 2013 at 19:31
  • $\begingroup$ @DavidZ I dunno, that makes it really hard to give an answer. If you look at the Good list, Bad list post where I mention the motivation, I think all of the bases (ie, all of the issues with traditional rec questions) are covered. $\endgroup$ Commented Nov 20, 2013 at 20:01
  • 2
    $\begingroup$ @DavidZ I would take serious issue with a policy that discouraged users from providing an answer with a single resource: it would mean you'd have to be an expert on the whole literature to speak up. If someone knows of a good book, I want to hear about it! $\endgroup$ Commented Nov 20, 2013 at 20:08
  • $\begingroup$ As I had understood it, the new policy was geared so that the description of each resource was as complete as possible, to avoid link-only answers for which one might as well go to Amazon. If someone knows of a good book but can't be bothered to write more than the title, then I don't see the point, but if they describe it appropriately, I don't see what the problem is with a single resource. $\endgroup$ Commented Nov 20, 2013 at 20:09
  • 1
    $\begingroup$ @Emilio No, you certainly don't have to be an expert on the whole literature. There's a huge difference between that and only being able to name a single book. The point, I thought, was that an answer should give an overview of the few most standard books on the topic. So someone who literally only knows one book on the topic shouldn't bother answering, but someone who knows the topic well enough to describe the top three or four could definitely contribute. (The matter of descriptions vs. links only is separate.) $\endgroup$
    – David Z
    Commented Nov 20, 2013 at 20:44
  • 2
    $\begingroup$ @DavidZ I strongly disagree. There's a huge number of scenarios where someone knows a book well but should not be discouraged from answering. Other books I know may already be in other answers. I may know of many books but only know one of them in enough detail to provide a worthy answer. The more specific the request is, the more likely such things are to happen. It isn't how many books you know that should matter, but how well you know them; it's the set of answers that matters. This is definitely what's documented in the current policy (paragraph below "How should I answer...?"). $\endgroup$ Commented Nov 20, 2013 at 21:07
  • $\begingroup$ (I would also like to know what the original architect, @Manishearth, had in mind when writing the policy.) $\endgroup$ Commented Nov 20, 2013 at 21:08
  • $\begingroup$ Ah, well I didn't notice that bit of the existing policy; if I had, I would have advocated clarifying/changing it to explicitly indicate that answers should be self-contained overviews. Certainly the example answers were chosen to indicate that. Also if other answers already include books you know, then nothing stops you from also mentioning them in your answer, which I would like to see be the recommended behavior. The point is that in the end, there should be one answer which fully addresses the question. $\endgroup$
    – David Z
    Commented Nov 20, 2013 at 21:57
  • $\begingroup$ @DavidZ I disagree that that should be the goal. It's definitely not what's documented at the moment. If you propose we change to that scheme, I would ask you to start a thread specifically about this point. $\endgroup$ Commented Nov 20, 2013 at 22:04


You must log in to answer this question.

Browse other questions tagged .