(skip to the "in short" section for the TL;DR)

One of the reasons of not allowing analysis of specific works is that the question and answer might not be useful to many people, potentially useful to the asker only, but that's not the case for all works. In fact, in the case of well known authors and works, that analysis can be interesting, useful, and contain a high academic value, while being important to musical practice and performance.

The idea is to allow analysis requests of specific music as long as it adheres to a specific criteria.

You can find more about that suggestion here: I think we should start allowing the analysis of specific musical works

To re-define the scope, we need to set that criteria.

Which kind of analysis should be allowed?

With the idea of usefulness in mind, I think and propose that the analysis should be harmonic at the very least and cover the work completely (not only selected parts).

Why "harmonic at the very least"?

  • It's the most common kind of analysis, so it will be useful to many people (assuming a well known work), and it's relevant to musical practice and performance.

  • Prevents questions with very reduced scope and usefulness and/or interest (even within the work), like "what's the time signature of this song?", or "what's the chord progression of this part of the song?", or "which are the chords of this song?". We could still allow these questions as long as they are part of a more complete analysis request (harmonic at the very least), but not if they are asked individually.

Analysis of which works should be allowed?

I'm of the idea of allowing analysis requests for any song. The community can express relevance, importance, and interest into the question via the voting system. We also leave the door open in case someone feels like answering a low interest question. Who knows? Maybe the analysis of that song in particular will be useful to someone else, even if it's not a big well known work at the moment of the question.

Another option is to allow analysis requests only about selected authors, works, eras, etc. We can use a combination of filters to make it as concise and complete as possible. I think common practice era is a must, but we can add authors and works from other periods and styles that are commonly analyzed or have the potential to be. Something like common practice era + Radiohead, Pink Floyd, ... (rock) + Coltrane, ... (Jazz) + etc.

This list of filters can be built somewhat objectively, based on popularity or relevance in the time and/or genre. It can be dynamic, it can accept requests.

The reason I'm proposing accepting all analysis requests without exceptions is that I think that filter list would be already implemented and expressed via the voting system. But who knows, maybe there's a good reason to still have a list of allowed analysis requests.

In short.

If we want to allow questions about the analysis of specific works we need to carve the scope. We need to discuss about:

  1. Which kind of analysis should be allowed?

  2. Analysis of which works should be allowed?

    • All of them?
    • Some of them? If so, which ones?

So, what do you think?

  • 1
    Disagree with harmonic analysis only. Any appropriate analysis should be fine.
    – topo morto
    Feb 23 '15 at 18:35
  • I can't see how anyone could justify or maintain a list of selected authors. Feb 12 '16 at 6:46

I think there should only be two requirements for an analysis question.

  1. That they have the sheet music(or some notation that shows all the melody and harmony) in the post or link to it.
  2. That there is an actual question about the analysis, not just "analyse this for me."

1 is important because without the score the analysis is still possible, but much more time consuming then it needs to be. 2 is important because there really isn't a question if you just want something analyzed for the sake of analyzing it. It kind of defeats the purpose of analysis.


  • 3
    I think #2 is the most important. I agree 100% Aug 5 '14 at 7:49
  • 1
    I completely agree with these points, @Dom, and would happily see them added to the guidelines. Aug 12 '14 at 21:51
  • 1
    Totally agree. These are very objective criteria. I should add that defining a corpus of works/genres/styles suitable for analysis is always vulnerable to taste matters, and highly idiosyncratic when it comes to deciding what is simple or complex, relevant or not, useful or futile. Voting and editing should work naturally to filter the general interest of the community on such questions.
    – SeuMenezes
    Aug 13 '14 at 6:16
  • 1
    Completely Agree! Aug 14 '14 at 13:18
  • Who says that music has to have melody and harmony - or even either?
    – topo morto
    Mar 3 '15 at 22:23
  • @topomorto then how would you analyze it if it has neither?
    – Dom Mod
    Mar 3 '15 at 22:28
  • I suppose it would depend on the question asked, but I hope we wouldn't cut ourselves off from answering questions about a polyrhythmic drum solo that might be hard to find a key signature for, or a section of throat singing where there's only one 'note' but the feature is the timbral changes...
    – topo morto
    Mar 3 '15 at 22:39
  • @topomorto There is rhythmic, form, and motif analysis that would still follow the same requirements as above that could be used on them. Regardless of what you're listening to you need to know what's going on before you can analyze it. Requiring the notation takes the transcription aspect out of it.
    – Dom Mod
    Mar 3 '15 at 22:56
  • Definitely good to require notation in cases where it can reasonably be given. I just don't think we want to be culturally biased against cases where notation is less easily applicable.
    – topo morto
    Mar 3 '15 at 23:08
  • Agreed that notation would make analysis much easier and eliminate transcription. But I believe it is possible that a particular work that represents a good example of some musical phenomena that analysis could explain - might not have readily available notation. I think that should be a guideline instead of a criteria ie: "If a transcription of the work used as an example in an analysis question is available - please proved a copy or link to a copy." Mar 4 '15 at 22:19
  • 1
    I agree about #2, but for #1, I'd expand that to notation, audio, or video being linked to. For my use case, as an aspiring pianist, I often hear things in songs that I cannot identify, and while I work on ear training, I get stuck sometimes. Being able to ask those types of questions would be greatly helpful to me, and I would open and read any question that asked this about any song in order to learn more for myself.
    – tarun
    Mar 5 '15 at 21:21

Based on my experience here and other SEs, I think the most important criteria relate to the quality of the question rather than the difficulty of the answer.

To explain: When somebody asks, “What’s the weird scale in this YouTube video,” it doesn’t really bother me that the answer might be very basic. What bugs me is that the only way to answer is by going to YouTube and listening to the whole thing and trying to figure out what the poster thought was weird about it. That’s the sort information that should go into the question. I actually enjoy answering easy questions, so I don’t care whether the answer is basic or tricky.

So I’m happy with any analysis questions no matter how simple. But please do some research effort, and give me enough written information in the question to answer it without needing to listen to random stuff on YouTube.

  • I think the term 'quality' cannot apply to questions, or there could be "stupid" questions; we all know that is impossible, it can be used axiomatically here. A question should be direct and specific, arguably meaningful (i.e. the answer, would help someone), but these are quantifiable things, not qualitative. Perhaps, though, you mean to say "the quality of the delivery of the question"? The question is the question, and there are good and bad ways to pose it. The nuance is about what a "question" really is, delivery or intent. I say 'quality' applies to the former, not the latter. Jun 11 '15 at 0:25
  • StackExchange judges questions for quality all the time. That’s what the voting is for. Jun 12 '15 at 0:22
  • You're right...I was getting a little esoteric there, trying to point out all the conflation of ideas that we smash into the word "quality" here...but from a high level you are right. Jun 12 '15 at 21:53

I think a discussion about criteria will get very abstract. And I certainly don't feel inclined to check the list of things people at one time thought should be allowed / disallowed whenever I feel like answering a question. Just let voting do its job.

  • 2
    +1, agreed. Interesting questions, with relevance beyond the scope of the piece or song, should receive more attention and votes from readers and answerers. Jul 31 '14 at 1:18
  • Yes, let voting do the job. The allow/disallow criteria on this site gets applied arbitrarily. It leads to closing/holding questions instead of editing and improving questions. Feb 12 '16 at 6:44
  1. All.
  2. All.

I'd concur that the sources should be readily available (and not subject to copyright takedown - brief examples from works under copyright should be feasible under fair use), and that the questioner should have made some attempt at research and should put up fairly detailed questions.

With these caveats, I'd have to say that restrictions are a bit counter-productive because one never knows when something from one kind of music will be relevant to another. I recall something about the conservative tonal composer Robert Simpson saying that he had learned more about large-scale tonal form from Arnold Schoenberg's (12-tone!) Piano Concerto than any contemporary symphony (and he really didn't like Schoenberg's music).

I'm answering this, though, primarily to bump the question up for further consideration because I think the whole question of analysis is a bit more important for illuminating musical praxis than the policies here assume. I joined here fairly recently, and was a little shocked to see that, when I ran across questions regarding analysis, they were being closed because "only of interest to the questioner"... on a site that is billed as "Music: Practice and Theory". I'm not sure you can have theory without analysis, and you definitely can't have analysis without analysis of specific works. In this regard, music theory is rather similar to the sciences - you need to deal with the particular in order to make a stab at generalising.

The danger of not allowing analysis is that theory can then (and does) tend to devolve into prescriptive "rules", and that is not really what theory is about.

Also germane to the question here is that we need to put aside objectivity as a criterion in favour of reasoned argument. Other than the sounds themselves and (when the music is notated) the notes on the paper, and (occasionally) facts of acoustics, psychoacoustics and sociology (and these latter rarely form a complete explanation for topics in music), there isn't that much that is objective about music (or any other art, for that matter). Reasoned argument based directly on elements of the music is entirely possible, however, and can and should be enforced by the mods.

  • I agree 100%. Not sure why there are not more up votes. I have a theory but wish not to postulate due to my suspicion that doing so might unwittingly imply that eloquence equates to relevance. But I will affirm the relevance of your post - in spite of ....... Mar 4 '15 at 22:38

You must log in to answer this question.

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