Should we authorize a select list of "Reference" questions

Question

It seems that daily we receive several "Where can I find references on ABC?" questions. It seems that there is an unmet need amongst candidates for members of this community/stack for reading lists on broad general topics in history. SO, along the lines of various early **list* questions on Stack Overflow. I am wondering if it might be worthwhile to authorize, as a community, and as a community service, a small number of questions listing some well-founded history approachable by the general public.

If so, I see a few issues that would have to be addressed:

1. How to determine the number of such questions, while still ensuring that this category does not overwhelm the site. One suggestion might be to allow one such question each month/week.

2. How to determine which question to open each as above? And who gets the final call.

Any thoughts on the merits of the idea, and on how it could be monitored and (if necessary) policed are welcome.

Answer 1

Yes, we should.

And we can make the answer(s) community wiki, or add a combined community wiki answer. We can sort out the technicalities later.

I am not formally a student of history, so I don't know how the historians decide which questions are objectively answerable (which seems to be the core philosophy of our SE). But sometimes I think the rules for on-topic-ness of our community are too restrictive for its own good.

Other sites (such as Math.SE, Theoretical CS.SE) allow a broad array of soft questions (examples 1, 2, 3, 4 — they're very easy to find because many SEs have a dedicated tag called soft-question), which can be a little subjective, and yet manage abiding to the core philosophy of SE, i.e. not getting converted into a discussion forum. Perhaps we can be a bit liberal in our scope, and be more helpful to a broader class of users.

Answer 2

Since its been a while, and thus I think a lot of people here might not be aware of it, I should probably put up my old Proposal for Accepting Reference List Requests.

The basic idea was to make a single accepted wiki answer, and allow other non-wiki answers with supporting detail from individual users, with the understanding that references in good answers are added (with no detail) to the accepted answer. In theory, the answers for specific references should end up appearing after the wiki answer in the order of user votes for them.

I believe we tried it once, with the result showing that at least one of the (very few) supporters didn't actually understand what I was proposing. I don't think it was tried again.

Answer 3

I am sympathetic to this proposal; I think it will solve a problem. I just don't quite understand how it will work. I'm going to try to summarize the proposal as I see it.

A. Proposal for reference questions

1. Reference questions are now permitted. The community will be relatively conservative about duplication - we want 1 reference question/topic, we do not want lots of closely related references. We acknowledge that there is an intrinsic problem with the scope of questions and will rely on moderation to converge. (However, B3 below may address this issue). NOTE: This does not imply pre-emptive questions. This merely grants people who are interested to ask questions like, "What are some sources to permit me to research X?".
2. Anyone can ask a reference question and site will rely on moderation to converge reference questions towards a loose standard.
3. Reference questions will have a distinctive tag (probably "reference question")
4. Reference questions will be community wiki so that anyone can edit.
5. Nobody makes the final call and nobody selects which questions; or to put it another way, the community moderates the site, like all other questions. I would suggest that as a community standard we be very liberal and accepting of reference requests.

B. Potential additions or alternatives

1. We don't do reference questions - we permit the tag wiki to contain a list of references for that tag. @called2voyage expresses a concern that tag wiki's may not drive enough traffic to make the result reliable. I'm not sure I understand the concern; the proposal would be better than the current reference system for H:SE. I'm not sure I understand the relationship between traffic and reliability (at the core, the problem may be that I don't understand how the term "reliability" is being used). In any case, I feel that I should amplify that I suggest the tag wiki because it seems to me to be the most intuitive place to look for resources, and it re-uses existing mechanisms. It requires minimal cultural adoption, and it scales well. I don't think that any of the potential solutions will be "reliable" and I don't anticipate any of them being high traffic.
2. In order to avoid debate/discussion, reference questions will have one answer that contains a list of all the canonical or commonly approved sources. Other answers can contain proposed resources. Any source that achieves a minimum score will be added to the canonical. This makes the debate explicit - up & down votes can be used to identify higher quality resources which are then moved to the "accepted" list. NOTE: this seems to have support and will probably be moved to A - the only reason I'm not moving it now is that it will break the references in comments.
3. Every reference question should have precisely two tags - the "reference question" tag and a topic tag. All references must be tightly relevant to that tag.
4. We don't delete answers with deprecated references - we annotate the answer with the reason why. (I don't remember which eminent computer scientist ?(Allan Kay?) suggested that the best reference for every topic is "Alice in Wonderland". That is great for humor, but lousy for a source question. We need a way to control both well meaning disagreement, and references with an agenda). (OP asked about how this would be policed; this is my answer, but I invite better answers.

C. Commentary

I remain confused about "select" in the question title and the phrase "small number" in the question body. I haven't seen a practical suggestion to make this a select list - every proposal I see is unbounded (with the exception of a time throttle). I'm hanging a lampshade on that problem and inviting others to address the issue. (might be that it has been addressed and I just don't understand the subtlety)

Answer 4

Building on Mark C. Wallace's answer, maybe we should have questions for some overarching categories that we decide in advance. The canonical answer can then be further subdivided as necessary. For example:

Question 1:

Resources for Time Periods

Note: Answer categories can be further subdivided by region.

Answer:

• Neolithic
  • Neolithic Ireland
• Bronze Age
• Medieval
  • Early Middle Ages
• Renaissance
• etc.

Question 2:

Resources for Specific Individuals

Answer:

• Philosophers
  • Plato
• Scientists
  • Einstein
• Generals
  • Patton

Question 3:

Resources for Specific Institutions

Answer:

• States
  • United Kingdom
• Companies
  • East India Company

Question 4:

Resources for Specific Events

Answer:

• The Black Death

Question 5:

Resources for Historiography

Answer:

• Comparative history

Question 6:

Resources for Archaeology

Answer:

• Remote sensing

---

The subdivisions aren't really important--they can be edited over time and refined since they will be in the community wiki answer. What is important is defining the overarching categories (the bold examples) in advance.

Answer 5

One idea for consideration: Put "good"/canonical references in tag wikis.

With it done this way, requests for references could be converted into (meta) requests to update that portion of a specific tag wiki.

The drawbacks I see to this are that we probably won't have specific tags for a lot of detailed requests, and its weird enough that it may be confusing to new users.

Also, there are reputation requirements (I believe overall, and tag-based) for tag wiki edits. However, that could possibly be viewed as a good thing.