# Book recommendations

There are quite a few soft questions asking for book recommendations. Reading some of these threads one might notice that there are answers containing several books at once. As such this might not raise any suspicion but in some cases one cannot help but wonder whether the recommender has actually read all the books they recommend.

Recommending a book to someone that one has not read oneself without explicitly declaring so makes no sense to me. If I ask someone for a recommendation I (obviously) expect them to have read what they recommend to me.

Do I have to lower my expectations when interacting with the MSE community? Or are there community members that share my view?

Be that as it may, I have trusted sources from which I can get book recommendations so that I don't have to depend on MSE for that. But what is still of slight concern is that one can gain a lot of reputation just by posting lots of recommendations (to maths one knows nothing about). In the worst case, this could be very misleading to unexperienced users and generally to people with little mathematical experience and this would be the opposite of helpful. As a community I think we are striving to help people rather than the opposite so perhaps this is an issue one might want to address.

-
We can hardly stop people from posting subpar answers. It would be reasonable for book recommendation questions to be converted to CW by the OP or by a moderator. – user53153 Dec 24 '12 at 17:13
Another point here is that the user may well be familiar with many books on the subject. A professor teaching an advanced undergraduate / beginning graduate course is likely to compare several textbooks in the process of choosing the right one[s] to use in the course. – user53153 Dec 24 '12 at 17:45
@PavelM Of course, "@"your second comment, but the majority of users here aren't lecturers. – Matt N. Dec 24 '12 at 18:36
@PavelM - to your second point: yes indeed; I do a lot of reading prior to selecting texts to use. And also, for those involved in administrative-curriculum-related decisions, and those of who write reviews of text, exposure to a variety of texts comes with the territory. – amWhy Dec 25 '12 at 0:33
But I do agree with Matt that perhaps more should be stated in a reference-recommendation-post as to why the recommendations are being made, and how or when or in what context one has encountered the text. – amWhy Dec 25 '12 at 0:35
@amWhy I agree, asimple list of books is a bad answer. – Michael Greinecker Dec 25 '12 at 15:11
Look at this post for example! How could anyone be familiar with this entire list! I do hate singling out any one post, but an example can be illustrative. In my own post to that answer, I included links to other posts, where there were more answers and recommendations made by others, to provide more input for the OP. – amWhy Jan 6 at 1:43

As somebody who has recently recommended books I have not read (I did acknowledge that I had not read or completely read the books), I would like to say the following:

I would justify my recommendations by pointing out that, as a graduate student in mathematics, I am surrounded by a culture in which a great many books are preceded by their reputations. I wouldn't recommend a book just because I saw it on Amazon, but I would recommend a book that multiple people have recommended to me. Usually, when I start reading a new book, I already have a good idea of what topics it covers, how thorough and elegant the explanations are, and how easy it is to read, even before laying my hands on a copy of the book.

-
 But what you say in "()" contradicts your first sentence: I complain about recommending books one has not read without acknowledging it. – Matt N. Dec 27 '12 at 17:13 @MattN. Yes, I've just removed the word "exactly." Implicit in your message was the idea that one cannot accurately recommend a book without having read it first. I'm sure many, perhaps most, MSE users are part of a culture that allows one to be reasonably familiar with many math books they have not read. So I wanted to share that perspective in this discussion. – Brett Frankel Dec 27 '12 at 17:23

Recommending a book to someone that one has not read oneself without explicitly declaring so makes no sense to me.

If you think it's reasonable to start reading a book because someone recommended it to you, why don't you think it's reasonable to recommend a book because someone recommended it to you? When you tell people things, do you generally make a habit of pointing out exactly where you obtained the information you tell them? If not, why do you want people to do this specifically for book recommendations?

I also agree with Pavel that you can't stop people from posting bad answers. If you don't like it when people recommend books without explaining why, then downvote those answers and maybe leave a comment. That's it.

-
Yes, if I tell people "things" I usually point out where I obtained the information or, if I can't remember I explicitly say that I can't remember. Otherwise what I say would not be a fact but merely a statement of opinion and hence completely irrelevant. As for recommendations: If I ask someone to recommend a book to me on $X$ the someone will (1.) know me well enough to know my taste (2.) know enough about $X$ to be a credible judge of the book they are going to recommend. You are of course right that one cannot stop people from posting bad answers. – Matt N. Dec 26 '12 at 16:54
(cont'd) But it seems much easier to spot a mistake in a calculation or proof than it is to notice that a list of books isn't suitable because for that one would have to have read the books. – Matt N. Dec 26 '12 at 16:55
I think that your first question misses the real point of Matt’s question. I am perfectly willing to make a second-hand recommendation, but I consider it dishonest to do so without noting that it is second-hand. Yes, I do make a point of telling people whether my knowledge is first-hand or not. – Brian M. Scott Dec 28 '12 at 7:51
@Brian: I just think Matt's declaration that the practice "makes no sense" is far too strong. At worst I would call the behavior Matt is calling out mildly negligent, but Matt is reacting to it as if it were immoral or something. – Qiaochu Yuan Dec 28 '12 at 8:09
I think that the makes no sense declaration is on a different axis from Matt’s real concern, which is the dishonesty of making a second-hand recommendation without acknowledging it as such. It’s a relatively minor dishonesty, but it is dishonest, and I don’t really see why anyone would engage in it save by accident $-$ under time pressure, for instance. – Brian M. Scott Dec 28 '12 at 8:16
"If you think it's reasonable to start reading a book because someone recommended it to you, why don't you think it's reasonable to recommend a book because someone recommended it to you?" - In the first case, I am trusting your judgement; in the second case, I am trusting your trust in someone else's judgement. – BlueRaja - Danny Pflughoeft Jan 9 at 21:20