(You can find the discussion about the criteria we could use to allow musical analysis here: Criteria for allowed analysis of specific music)

Analysis requests of specific works (time signature, harmony, etc) are currently not allowed in Music SE. I understand why: "Questions on identifying (or finding) a particular song, genre, instrument, etc. are off-topic since they are rarely useful to future readers."

I agree with this, but I also think this not necessarily applies to all works, and that we are letting that little con completely shut down a very big and interesting pro: the harmonic, objective, analysis of well-known musical works and/or authors, that would be of the interest of a huge amount of people that would find it useful. Analysis of big songs, by big authors. (I know, what is "big", right? I think we can draw that line, bear with me)

We can either:

1) Allow the analysis of all musical works. No exceptions. We will get some questions with a very low interest amount, but what's the worst that could happen? Some questions not being answered? I think that's a very low "price" to pay for opening the possibility of analysis of works and/or composers that carry a lot of interest, that have the potential to generate great questions and answers.

The analysis of less-known works will always carry the potential of being useful to someone else anyway.

2) Allow the analysis of some musical works, based in a criteria set by the community. If we think opening the valve completely is not a good idea, maybe we can open it just a little. We can set some rules. Here are some examples:

  • Create a community question once a week (or month, or every two days, or whatever) about the analysis of a specific work or author. We can vote and discuss the specifics of the question (which work, which author, etc).

  • Allow analysis requests from only selected authors and/or eras and/or any other filter. Maybe allow common practice works only, or maybe only from specific authors, or whatever you can think of. Again, we can set and discuss the specifics here, in the meta.

  • Allow analysis requests only from users above X reputation (I know this is out of pattern with other SE sites, but it might not be a bad idea to implement it on this scenario).

This type of question will likely end up being on topic in the Music Fan SE Proposal, which is close to reach the commitment stage. I think it makes much more sense to have them here.

What do you think?

  • Or keep it the same and not allow analysis of specific work. Trust me I really do like analyzing pieces of music, but it doesn't feel at home here especially since an extensive analysis of even a simple song can take quite a bit of time. Analyzing a song or transcribing a song is an important and powerful skill for a musician, but it's important that it is learned and not done for someone.
    – Dom Mod
    Commented Jul 29, 2014 at 1:26
  • 7
    @Dom What's wrong with extensive answers? What's wrong with time consuming answers? I personally have made giant answers and it didn't seem to be an issue, on the contrary, they were highly upvoted. The analysis in question is useful in one than more way, and in your specific example it can be useful for people that already completed the analysis and want to compare it with others and/or check for any mistakes. Why do you think we shouldn't allow any kind of analysis of specific works? Which are your reasons? Maybe I'm missing something here. Commented Jul 29, 2014 at 2:09
  • @Dom Maybe you can put those reasons in an answer to this question (along with your extended opinion, maybe?), so we can discuss them more extensively and specifically. Commented Jul 29, 2014 at 2:16
  • What about requiring some work from the questioner? A low-quality, low-effort post my just ask for the harmonic/rhythmic analysis of a song; a better question would basically be "here is the work I've done. I'm a little stuck on the section from measures 35-40", for example. Commented Jul 29, 2014 at 19:15
  • 4
    @TheChaz2.0 I don't see why previous research should be required. Why overly complicate something so simple as a Q and A? What is X? It is Y. What's the harmonic analysis of X? This is the harmonic analysis of X. I never understood that "previous effort and research" requisite some of us sometimes try to push into the site. It might just be me, but I think the existence of previous research is unimportant to the actual question and answer. Commented Jul 29, 2014 at 19:21
  • 2
    This issue seems to parallel that of homework questions over on math.se, where I was a very active member for some years. The idea (which is supported by policy over on m.se) is that requiring some effort from the OP automatically raises the bar. We don't have to rate which songs/topics/authors are worthy of not being closed; the questioners filter themselves by the effort they put in. Commented Jul 29, 2014 at 19:25
  • Of course, we could just let any questions on the site, and try to be answers.yahoo.2.0 Commented Jul 29, 2014 at 19:26
  • 2
    @TheChaz2.0 I do see where you are coming from though. Maybe doing someone's homework, or something like that. I agree, that's lazy. With that in mind, in the great scheme of things, I think the relevance, importance, and interest should be focused in the question by its own merits. I see it as we are one of those sites where people perform research, where people find answers. That's why asking for previous research does not make sense to me in the context of this site. Commented Jul 29, 2014 at 19:33
  • 1
    @TheChaz2.0 Maybe you are over generalizing. Math SE does welcome and includes several popular analysis of theorems and other works from popular Mathematicians. The homework questions that are not welcome are the ones like "solve me this equation", which here would translate to "which is this interval" or something like that. I see the analysis of musical works as the analysis of theorems, like this one (112 upvotes): math.stackexchange.com/questions/675522/… Commented Jul 29, 2014 at 19:37
  • Just to clarify - I don't mean previous, unrelated research; I'm talking about showing the work that you have done (even if it's just 5 minutes' worth!) towards the particular song's analysis. Commented Jul 29, 2014 at 19:39
  • 1
    @TheChaz2.0 I don't agree, but I'm interested in what the community thinks about requiring previous research in questions. Maybe you want to present your points as an answer to this, or continue the discussion in chat? (having extended discussions in the comment section is very uncomfortable) Commented Jul 29, 2014 at 19:43
  • I’m not sure whether this new question asking to identify scales in several songs is evidence for or against this idea. What would you want to see to make a question like that answerable? Commented Aug 2, 2014 at 1:46
  • @BraddSzonye Seems that most people think that we should let the voting system do its work as an alternative to closing everything related to analysis of works. The only problems I see with that question are 1) The title should include composer's name and 2) It should be split in other questions, one for each piece. Commented Aug 2, 2014 at 1:52
  • 2
    I like this idea and agree, music p&t should allow something but ensure the line is there. We respond to many different questions about how becoming a better musician can involve transcribing and analyzing solos. There are questions like music.stackexchange.com/questions/29715/… which are about specific songs but lead to answers that specify general techniques.
    – tarun
    Commented Feb 12, 2015 at 14:13
  • 2
    I genuinely wonder if there's an inherent problem in requiring research & effort: different people with different levels of skill & experience can put in equal amounts of research and effort, & they will achieve very different amounts of progress and insight into their question. I used to spend some time on Physics.SE (maybe Math.SE has the same quality--I haven't spent much time there), & there was a tendency to immediately label simple questions as being poorly researched. I think it's possible for simple questions to have been well-researched but for the asker to have made little progress.
    – jdjazz
    Commented Apr 30, 2020 at 0:27

6 Answers 6


I personally would like to read (and ask) interesting questions about the analysis of specific works and composers.

However, I am also strongly opposed to basic identification questions like “What’s the key of this song I found on YouTube?” The recent question about time signatures was a bit better, as it actually showed some research effort, although I still encouraged the poster to include more information about the song in the question itself, rather than relying on a YouTube link. If the video gets taken down, we don’t even know who the composer is.

Thus, I am in favor of encouraging analysis questions if and only if the questions also follow the usual Stack Exchange criteria like showing research effort and putting all of the necessary information into the question itself. Those are my recommended criteria.

  • I agree wholeheartedly with your first sentence. It's music theory at its best, IMHO!
    – jdjazz
    Commented Apr 30, 2020 at 0:28

This is a question we need to take up and resolve as a community, because it hasn't been going away.

We recently closed and reopened a question that effectively asked "What is this compositional rhythmic technique, and how does it work?", and referenced a specific passage of a specific work (see How does one analyze tricky time shifts?). Personally, I would think that is precisely the sort of question that would work on a Q & A forum on music performance and theory - a general question like "What is hemiola and how does it work?" won't cut it because, like most novices, OP lacked the terminology and was unable to recognise precisely what he was hearing. What he was hearing, however, was a technique of very wide utility in very wide usage, and hence of general interest.

My own view on the matter is based on the following premises:

  1. Theory that purports to describe what goes on in music is not really possible without analysis. Theory that lacks analysis would be, pretty much by definition, prescriptive. I've run across a few that come close to this latter state of affairs; they were, shall we say, less than convincing.
  2. You cannot really teach analysis without something specific to analyse. Analytic techniques that purport to work with everything don't - one size does not fit all, primarily because most analytic techniques are reductive in nature, and thus cannot cover everything of interest. (I'm reminded of the story of Schoenberg looking over Schenker's analysis of, I believe, one of Schubert's symphonies, and saying, "Where are my favourite passages? Oh, there they are, in all those little notes!")
  3. The corollary to 1 & 2 is that a lot of questions will require specific examples, and it is in many ways a moot point whether they are provided with the question or by one of the answers (or both).

Note also that the question I referenced above never really changed before it was reopened; a little "syntactic sugar" was added to make it acceptable, and that shouldn't have been necessary in the first place. The OP asked a fairly precise question, and was genuinely interested in how the technique works. When how something works becomes a focus, you will generally get answers of general interest. It may be wise in cases like this, however, to add a tag or two to make the question and answers properly searchable when similar questions come up.

I'm generally in favour of keeping the range of questions open, because it is impossible to know when something in one kind of music may provide pointers for a very different style or genre. For example, I'm learning more about jazz and pop theory here than I have from friends in those fields - not all practitioners are good at describing what they do. This is useful to me, because my classical compositions frequently have a faint whiff of prog and jazz in the background (the stuff I grew up with, eh?).

Broken windows should not be a problem. We do not need to abandon "of interest only to the person asking the question." We just need to cut the tie between that reason for closing and the analysis of specific works. The questions that we worry about cluttering the site, questions such as "What chord is he playing here?" or "What's the time signature here?", aren't analysis, not when a cursory read of any associated sheet music can give the answer.

  • 1
    Agreed. The only change needed is that when there's doubt, people should vote to keep open instead of closing. There must be something very pleasing with voting to close. I think we need to make the closing reasons very precise, in order for them not to be misused. Or remove a lot of them and have a case to case basis... Commented Feb 9, 2015 at 22:42
  • 3
    Not sure in some cases - asking for product recommendations is out-of-place, so I rarely have a problem with that. The policies as they pertain to analysis and musicology strike me as wrong-headed. The training of a whole generation of very well-known American composers was most strongly influenced by a teacher who was not a composer or theorist, but a specialist in déchiffrage, the analysis of specific works. I am referring to Nadia Boulanger... There is a lot of "Eureka!" in analysis that even crowd-sourcing (alone) can't provide.
    – user16935
    Commented Feb 9, 2015 at 23:52
  • 1
    Agreed, @ToddWilcox, but there are more than a few ways of asking or implying that. It's maybe wise just to make sure we read carefully before we vote to close, I think.
    – user16935
    Commented May 20, 2016 at 18:24
  • @ToddWilcox, that comment appears to be based on an unkind stereotype of a group of people and an entire genre genre. I can't imagine someone who likes indie music will feel welcomed or feel motivated to stay on this site after reading that comment.
    – jdjazz
    Commented Apr 30, 2020 at 0:31

I can see this working, and I'd support it, however, I think the criteria may be very subjective. Analysis of major composers works (eg Beethoven) I would imagine would probably be accepted, and is likely to be of interest to a large number of music students and hobbyists.

But what about the musical analysis of, say, Baby Got Back by Sir Mixalot - is that likely to tie in with a music syllabus, or general interest in music theory? It might, but I find it hard to see it.

So I'd be tempted to go with your option 2. As regards your sub options, I don't think it is right to limit it to high rep users - we may have new members of the community studying music theory and who happen on this site, so we should encourage them.

Which means we'd need to come up with a straw man for the criteria that could work. You might want to raise a separate question on suitable criteria for objective musical analysis.

I don't think this will be easy, and unfortunately I don't have the formal music theory training to be able to identify the key criteria, but I think if you can build it, and enough members of the community agree with you, then the moderator team will support this. If it works well, we edit the faq and help pages accordingly. If it doesn't work, what have we lost aside from a bit of extra effort?

  • Yes, the criteria definitely will be subjective, but we can carve it to focus on works that are normally the target of such analysis, which can be done objectively; those works are the ones that are most likely to be useful to others. Commented Jul 29, 2014 at 16:04
  • 1
    I have trouble with option #2. The basic justification seems to be that we'll get less crap. But if someone asks an excellent question that doesn't fit the criteria, they'll be sent away, and it's really quite arbitrary. I oppose that. As an off-the-cuff estimate, I'd say we get a dozen "What scale is this?" questions for every interesting one; thus I am not very comfortable with option #1, either.
    – user28
    Commented Aug 2, 2014 at 15:18
  • So @MatthewRead, are the criteria "about right" already, where analysis of specific works is generally off-topic, but the community as a whole can be trusted to keep open questions that have interest beyond the scope of the specific piece/song? Commented Aug 2, 2014 at 15:26
  • 1
    @BobBroadley I think so. I've definitely seen some great musicology questions accepted or re-opened in the past couple months, though an example escapes me at the moment. There are very few times when I disagree with a re-opening, but in those cases it's my job to accept the community decision unless there is a serious issue beyond the topic of the question. It might be possible to determine the factors that make such questions great and build our guidelines around them -- but I think it has more to do with the nature of the question than who the composer is, as proposed here.
    – user28
    Commented Aug 2, 2014 at 15:36
  • @MatthewRead Excellent questions are already being closed because they are about analysis of specific works. That's a flaw of the current system too, something a new system would improve. And re-opening wouldn't be exclusive to the current system, it can work with off-list works too. The current system is at flawed as it gets, as analysis of specific works questions are not allowed at all, even if they are important, interesting, and relevant. Commented Aug 2, 2014 at 19:11
  • @MatthewRead What is very clear is that the community wants to move away from that overly restrictive approach. What's your suggestion? Commented Aug 2, 2014 at 19:23
  • 1
    @JCPedroza It might be possible to determine the factors that make such questions great and build our guidelines around them. I do not have anything concrete at the moment. I also don't think it is so clear -- there are several members vocally in favor of this, several members who are continuing to close and resist re-opening of such questions, and a whole lot of discussion (which is great).
    – user28
    Commented Aug 3, 2014 at 20:25
  • 3
    @MatthewRead It's very possible that they are closing because this subject is still in early discussion stages. There is nothing official yet. That also means that it is very possible that they don't even know about this discussion yet. I don't think it's fair or useful to use close votes as argument in this context. It's obvious that we still have things to do before implementing that change. Commented Aug 3, 2014 at 21:48
  • Sorry - just now joining the discussion. As JC points out - I did not know about it until recently. But I am strongly opposed to building a filter to exclude certain works or only allow certain type works. Read next to last paragraph in @Patrx2 answer. Also, I am not in favor of discrimination based on genre' era of origin, or any other arbitrary criteria that might exclude otherwise great opportunities to learn through analysis. Commented Mar 4, 2015 at 21:40

Broadly, I share @DrMayhem's point of view. And I can see why "musical analysis" and "musicology" questions and answers have been difficult to reconcile with the SE format thus far. But, personally, I would be happy with @JCPedroza's Option 1.

Musical analysis and musicology are always going to be more subjective than purely factual answers, but it would be great if we could trust the "system" to promote or demote interesting and less-interesting questions, respectively. Questions considered to have an interest beyond simply the scope of the piece or song in question, will get voted up by the community, and hopefully receive attention from the wealth of expert users visiting the site. The kind of questions we already try to discourage ("What are the chords in this song?", "What instruments are used in this piece?", "What is the sound at 35 seconds?", will no doubt: be voted down; receive less attention; or at the very least, not be voted up.

Equally, answers that draw out interesting information beyond the scope of the piece/song in question, will hopefully create interest within the community, and so receive upvotes. Those that simply give a "homework" answer ("It goes E A D G.", "The instruments enter in this order: guitar, bass, drums, nose-flute.", "The guitar uses a flanger."), will again; be voted down; receive less attention; or not be up voted.

These more subjective questions and answers may not be ideal for the SE format, but they do (already!) raise us above a mere wiki.

In the six months I've been using this site, I have noticed that the QPD has started to increase, and so we are not "desperate" for questions. And generally it seems that it is pretty obvious which questions are off-topic. It would be great, though, if we could embrace analytical and musicological questions. I am biased, as these subjects are of great interest to me, but I reckon they are pretty important aspects of musical practice and performance.

Having said all this, I largely like the way the site runs already, and I do feel that I would need to have been using the site for a little longer, to be able to make too strong a case for any significant change of focus.

  • 2
    We do not want a class of questions that are voted down and ignored. Valid questions that are treated that way will sometimes crop up while still being on-topic and permitted, but it should be a rare exception. You can't trust the system to take care of such questions because they cannot be closed and thus are not deleted; they sit around, with ugly downvotes, making new visitors think our site is full of crap (or alternately, think that it's OK to fill the site with more of them; this is referred to as the "broken windows" problem).
    – user28
    Commented Aug 2, 2014 at 15:13
  • 1
    Right, I understand. So is opening new questions the best way forward for analytical questions, as per this meta-post: meta.music.stackexchange.com/a/851/9198. BTW, not heard "broken windows problem" before, I like the analogy! Commented Aug 2, 2014 at 15:15
  • I definitely favor those sorts of questions. Teach a man to fish, and all that. Here is more detail on that analogy if you are interested :)
    – user28
    Commented Aug 2, 2014 at 15:25
  • Great, I like this idea too… And then, presumably, if little value is then seen in the "specific" question, it can be closed as a duplicate, right? Commented Aug 2, 2014 at 15:28
  • Hmm. That is something I think we might need to discuss further. I'd absolutely want to point the asker to the "general" question, but I think close-as-duplicate is meant for closely matching questions rather than "You will be able to answer your question by learning the techniques posted as answers to this other question".
    – user28
    Commented Aug 2, 2014 at 15:31
  • 1
    So, just closed as "off-topic", as we do now (easy!) Commented Aug 2, 2014 at 15:33
  • @MatthewRead I don't understand. Seems that you guys are suggesting to open new questions, but wouldn't that new question still be about analysis anyway? Exactly what are you guys suggesting to do with analysis questions? Commented Aug 2, 2014 at 18:58
  • Hi JC, I think the idea is to open an analytical question that is less specific than the original question, where the original question is very specific. This means that it isn't necessary to edit the content of the original question to make it more general, and so "on-topic". Commented Aug 2, 2014 at 19:02
  • @JCPedroza Questions like "How can I identify scales?" have always been on-topic. Analysis, as a whole, is not off-topic.
    – user28
    Commented Aug 3, 2014 at 20:23
  • @BobBroadley But that doesn't solve the problem. Am I missing something? What we want is to allow the analysis of specific works, which is inherently and inevitably specific. How do you ask a question like that without "being too specific"? Commented Aug 3, 2014 at 21:56
  • Hello again, JC. Personally, I'm happy to see questions about analysing specific pieces/songs. This is something of real interest to me. I would be happy for analysis of specific music to become "on-topic". If the guidelines are changed, though, I would suggest that something is mentioned in the wording to encourage questions asking for actual analysis and not just questions asking for us to "work music out for them". Commented Aug 9, 2014 at 9:45

"Questions on identifying (or finding) a particular song, genre, instrument, etc. are off-topic since they are rarely useful to future readers."

I've said it before, and I haven't changed my mind: I believe that the assumption that this is "rarely useful" is false. I agree analysis should be allowed and is in fact very useful.

"1) Allow the analysis of all musical works. No exceptions."

I would be in favor of this. Low quality questions simply won't get answered. Low quality answers simply won't get upvoted. Voila.

"2) Allow the analysis of some musical works, based in a criteria set by the community."

Please, no. This will just result in question voted for closing being discussed over and over on which criteria were or weren't met, to what extent etc and so on.


Music theory is a developing field, but probably not developing fast enough to make a healthy stream of non-duplicate questions likely. If we don't allow specific works to be referenced, Isn't the site destined to die?

It's not like Stack Overflow where there are dozens of new technologies every year to ask questions about. (There are lots of new products, but I find the "how do I do this in this program?" questions are no more likely to be of interest to other readers than those analysing specific works, and we presumably don't want this to become a tech support site).

  • I am very new to this site and quite possibly mistaken, but would appreciate some explanation for the downvote. Commented Feb 23, 2015 at 12:07
  • 1
    I think you are underestimating the depth of the field of music theory. We haven't so much as scratched the surface of it here, and we cover a lot more than just theory (which in turn is a lot more than just analysis). In fact, specific works are much more of a dead end than you think. It's like asking about Revision 3 of a specific source file. Programming languages and technologies are tools for creating such code, and are infinitely open-ended, much like musical tools and techniques.
    – user28
    Commented Feb 24, 2015 at 19:57
  • 1
    You're saying that without an (essentially) infinite set whose members have finite properties, we're doomed, but we're already working with other infinite sets with infinite properties and infinite applications.
    – user28
    Commented Feb 24, 2015 at 19:59
  • @MatthewRead Thanks for the thoughts. I don't quite resonate with that 'set' breakdown of the issue. As for asking about a specific source file - that's a key part of what SO does, allowing people to ask about specific code samples. Commented Feb 24, 2015 at 20:44
  • In my comments above I am only addressing the content of your answer, not trying to make an argument for or against these questions. What SO does is ultimately irrelevant to our scope, though.
    – user28
    Commented Feb 24, 2015 at 20:54
  • 1
    @MatthewRead My question was prompted by the fact that after 1400 days in beta, all the site stats are Excellent apart from questions per day, which is just OK. If it's true that we haven't scratched the surface of music theory in over 5000 questions so far, then at the rate of 7.5 questions per day it seems that we're not going to make significant inroads in my lifetime (and I'm not that old!). I guess my thought is that some specific analysis questions could actually help with the scratching! Commented Feb 24, 2015 at 20:55
  • 1
    @MatthewRead So far, after 7 months of this thread, seems that the majority of the community is in favor of allowing the analysis of specific musical works. What else is needed to be done for this change to go live? Commented Feb 25, 2015 at 18:20
  • 1
    @JCPedroza Propose specific criteria for determining which questions we do and don't allow. All of the 4 top answers don't want to allow basic "what chord is this" analysis (although one asserts those "aren't analysis" which is unhelpful for distinguishing them).
    – user28
    Commented Feb 25, 2015 at 19:12
  • so could @JCPedroza start a new meta thread announcing that analysis of specific musical works is now on topic and what consensus has been built - and then the discussion under that would fill out the details? Commented Feb 25, 2015 at 21:07

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .