Why is book recommendation considered off topic in History SE? It seems so snobbish. Many most vibrant SE's like Mathematics have book recommendations. Can we have too please? It encourages personal reading, the exact venue from which all answers from wherever, to wherever come. So if we stimulate further interest or knowledge, what's wrong guys?
I could see where it might seem weird at first blush, but I'm not real sure how it comes off as "snobbish". Presumably everyone here loves history books and reads rather a lot of them.
The issue is that such questions just don't fit our format. Questions here need to be capable of having a single answer worthy of receiving an "Accept". If someone asks "What are good books on Topic X", and I answer with 3 good books and someone else answers with 3 other good books, which answer should be accepted? How do you judge one better than the others?
I suppose we could try to answer such questions with a single Wiki answer (what does everyone else think?), but only convention would keep someone from spoiling the system by adding extra answers, and contributors to a Wiki answer can not receive any reputation reward for their good work. The StackExchange software isn't set up well to help us out there.
I went over to the Math exchange to see how their system works. The results seem kind of mixed. On the one hand, they do seem to be accepting book recommendation questions without their entire stack burning down or anything. On the other hand, I can't say I find any of the results particularly useful. Recommendation answers typically seem to get no more than one or two upvotes, and acceptances seem random on the rare occasions where they happen at all.
A great way to understand waht good book / reference requests look like is to review the questions asked on Mathematics.SE, under the
reference-request tag. Even the book-recommendations tag has questions that are specific in their learning aim.
Above all else, questions are specific to a facet of the topic. They do not engage broad enquiry or foundations learning.
Other than the problem of selecting an accepted answer, another issue with book recommendation requests is that they often tend to be quite broad because the OP is looking for an introduction or overview of a period/event rather than having a specific question.
When a request for a book recommendation is quite specific then it can usually be rewritten as a question (to match the SE format). The best responses should not only answer the question but should also include references and sources. So History SE gets a suitable question and answer and the OP will get some reading recommendations as a by-product.
However, when the request is, for example, for an overview of the Middle Ages, it's not possible to render it as a suitable question because it would be too broad for the History SE format.
The problem with book requests is that they are not asking a question. They are basically asking you to do research for you. Petty research, but nevertheless research. Also, speaking of pettiness, 95% of such requests are trivial to fulfill. OPs should be finding their own books, which is easy to do in most cases. History.SE is for hard questions that cannot be answered by trivial Google searches. Most book searches fall into that category.
That said, if a question involves a very rare book or source, that an easy web search would not reveal, then I think it is legitimate, but as above this is a 1% situation.