# Canonical book recommendations versus Finding Things Questions

As a community, we've come down pretty clearly against "Finding things questions" (FTQ). That said, there are some edge cases that deserve additional consideration by the community.

One case for us to consider is "canonical book recommendations." A canonical book is essentially a well regarded book that most in the field would consider to be a definitive source on a particular subject. Every engineering field has them, and some subjects have more than one. If I weren't lazy, I'd walk over to my library to cite a couple of examples.

In general, StackExchange is pretty much against book recommendations. But the technology focus of the "Big Three" sites may be just as much to blame for that stance. A few SE sites have managed to be able to allow those types of questions without them getting out of hand.

It's worth noting that engineering technologies do not turn over anywhere near as fast as computing technologies. And one of the principle complaints against book recommendations is that they become dated very quickly.

Some questions then:

• Do we want to allow questions about canonical engineering references?

If so...

• How do we control for the "I didn't bother attempting any research at all, what's the canonical book for foo" type questions?

• How do we handle situations where there isn't a canonical reference?

• How do we handle situations where there are multiple canonical references but a given answerer may only know of one?

• How do we handle cases where a canonical reference kind of covers the question but not completely because of XYZ and ... ?

References

Book Recommendation Questions

• I am interpreting this question as not meaning to include codes/standards as I discussed in this answer to the Finding Things Question? Seems like Roark's Formulas for Stress and Strain or Bowles' Foundation Analysis and Design might be the kinds of "canonical" sources to which you are referring? – Rick supports Monica Feb 3 '15 at 4:40
• One possible problem is that even canonical resources can become outdated and new books come out all the time. But the votes are after they are cast not so easily redone. I fear such a list would always be kind of outdated. Nevertheless one could maybe do it for the big areas in a community wiki kind of fashion? – Trilarion Feb 3 '15 at 17:30
• For an example of a site that's managed to get a good workflow with book questions, see the resource-recommendations policy at Physics.SE. Some of the better Q&A threads are here. – E.P. Feb 10 '15 at 22:28

• As I mentioned in my answer to the FTQ question, I think that this should be limited to clearly authoritative references such as engineering codes and standards.

• I also mentioned engineering societies because I think that something along the lines of "I'm a member of ASME in the US but I'm moving to Germany, is there an equivalent organization there?" is a valid question.
• I'm not entirely convinced about canonical textbooks. I think that this could be very useful in theory but it's a slippery slope when you start having diverse opinions from diverse parts of the world. I'm not sure if there are any truly universal textbooks out there...
• I would vote to close an unresearched "canonical book" question.

• I would upvote the first person to answer saying that there is no canonical reference. ;)
• An answer missing additional canonical references should be edited to add the others if there are maybe 2-3 such references... but that's a slippery slope. (See above) I'm very wary of "canonical" references.
• I think the limitations of a semi-relevant "canonical" reference should be very carefully explained. But, again, slippery slope and I'm not in favor.

Bottom line: I think that non-authoritative references would quickly become a muddy discussion of "which textbook is best" which would probably devolve into "which textbook did I use in undergrad/grad school." I don't that would be beneficial for this site.

• I see the reference to "canonical books" as not really including textbooks. There definitely are some similarities, but I would think a canonical book is more like a comprehensive reference guide. Without already knowing the principles, it might not do much good, it's more tables and formulas for various general cases that can be used as a basis for calculations. And there will always be opinion, but that's why we limit to only the serious consensus materials. – Trevor Archibald Feb 3 '15 at 4:11
• I understand what you mean by consensus (in the profession), but how do you have consensus in the SE Q&A format when there is only one questioner (who chooses the accepted answer) and one answerer? I suppose the voting system can take care of this, but the "wrong" answer can easily end up as the accepted answer. – Rick supports Monica Feb 3 '15 at 4:36
• That's why my proposal is to create a list that the community votes on that can be kept in the wiki or as a meta post that can be referenced. When someone asks for suggestions on reference books, we can direct them to that list. – Trevor Archibald Feb 3 '15 at 14:28
• In theory, such a list could be very useful. But this is the Internet and many people have many different opinions. If there was a community consensus to make such a list, we would need to be very careful to avoid conflict while preventing the list from getting so long as to be useless. I'm not saying it's impossible, just that it would be a very non-trivial undertaking. – Olek Wojnar Feb 4 '15 at 1:59

I'm not sure if this is an accepted practice on SE sites (or if that even matters) and how we would best go about implementing this, but I think, if these references are truly canonical, we could simply create a list of them. I'd expect it to be a good size, but not truly unmanageable. I wouldn't expect these to be very specific applications, so for each discipline (i.e. degree-level classifications) there would probably be not more than a handful.

For example, I got my degree in mechanical engineering, and I'm familiar with Roark's Formulas for Stress and Strain, Shigley's Mechanical Engineering Design, and Marks' Standard Handbook (plus one other I saw on my boss' desk that I can't remember the name of). But between those 3/4, mechanical design is covered pretty extensively and other mechanical engineering fields are covered at least a bit.

I wouldn't expect the other disciplines to have many more comprehensive books than ME. So if we could source a list from the community and then post it somewhere that we can easily find it and link to when questions of this nature arise, we can easily close them and hopefully quickly redirect people to the list to eliminate most of them down the road.

• Could or should such a list be kept in the community wiki? – Fred Feb 3 '15 at 11:27

I don't share the obsession over keeping out 'bad' questions. If it's a question where an answer would be genuinly helpful to the asker and it's a real problem arising at work or in an interesting hobby project, it's ok in my book. If a question seems poorly researched, my answer would probably miss the askers problem so I don't bother.
Look at eng-tips: probably 80% of the questions asked there would be closed or heavily edited here, yet there's a huge base of users, some of them highly knowledgable. So I don't believe in a broken window theory.
In this light, here's the approach I'd favor. I assume many here will disagree, even if so you can hopefully find parts to adopt and adapt.

How do we control for the "I didn't bother attempting any research at all, what's the canonical book for foo" type questions?

Depends on the question I guess. Is it a question someone conceivably might ask at work: "I'm asked to check wether our mineral slurry pumps are suitable for applications in waste processing. Where can I get up to speed on the specific (material etc ...) challenges in waste processing?" ... I'd see it as a good question. "What's a textbook on waste processing" on the other hand ... my answer would probably miss the askers problem anyway, so why bother.

How do we handle situations where there isn't a canonical reference?

The asker, and in all likelyhood the person(s) voting to close, won't kno wethere there's a canonical reference, or several, or wether the field is still strongly in motion. Why not leave it to an answerer to say: "Any recent textbook will do, there's no canonical source." This would actually be a hepful answer for some.

How do we handle situations where there are multiple canonical references but a given answerer may only know of one?

Again, the asker will (hopefully) have mentioned a specific problem he wants to understand. An answer should include why this source is the most relevant to the askers question. At the end we have several answers with explanation and we have saved the asker some time in research.

How do we handle cases where a canonical reference kind of covers the question but not completely because of XYZ and ... ?

Say so in the answer. This assumes that the asker wrote more than "textbook XYZ plz", but we assume that anyway.