I am mostly referring to composers writing complex music, for example, classic composers. In particular, I am referring to change of harmony, and change of key. Let's take for example a fugue for Bach, where the harmony is changing almost for every beat and there are very frequent modulations. How much are this modulations and change of chords "conscious" and how much they just follow the strongest melody line and naturally come out of them, thanks to the compose ear? When Bach or Mozart go to (let's say) I to [IV of IV] and [V of IV], when he is using a Napolean chord and so on... do they actually know they are doing that while composing? Do they actually want to do that? Or are they just chasing a sound that, from our analysis, resulted in being that?

It is obvious that at least some of them are at least partially aware of this. Mozart's letters to his father show his awareness on what modulation to use for obtaining a specific effect. However, I wonder: is this always the case? Is it possible that Mozart was just "reading" what he naturally felt right? Meaning: maybe he was not aware that he had to do that modulation for obtaining that effect, but he tried it and it worked and then analysed it for referring to his father.

And, this is only Mozart. We barely have any letter from Beethoven that talks about the structural harmony of his music (maybe no letter at all about this?).

I am referring not only to great classical composers but also to modern composers. They don't need to be classical, but of course I am interested in cases with complex harmony and modulations, so pop songs, for example, won't do.

I appreciate both personal experiences as composers or more detailed account from the literature of the past/present.

EDIT: I am still mostly unsatisfied with the answers, and I think that my question needs clarification. When composing music, does a composer go like: "Uhm, here I should put the V chord in second inversion as is the thing that follows best the II chord" (I am making this up) or like "ah! This sounds very good in this position" [and only after that maybe analysing from a harmonic point of view]. Which one would they mostly use?

EDIT2: This question is put on hold because "answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise". I think is unjustified. I am asking about the mental process of composers while they compose (referring to the most famous composers in order to assume their proficiency). Literature on this topic does not exist, apparently, but it could, either through the huge amount of letters that composer wrote during their lifetime, or through personal experience of living composers. I think is important to know whether this literature exists or not. If every composer uses a different composition style, then it would be good to know. From the answers, it seems that there is no consent about that. Some people believe that of course they are totally aware of the harmony, someone else says that they may be, maybe no. Some of this answers are unjustified and based on personal opinion. However, is totally possible to obtain an answer that is indeed based on facts: e.g. a big literature review on the letters of the great composers (or the first-hand experience of the less famous one), showing that things go mostly on one side or the other. Someone was writing a big answer on this. Why prevent potential factual answers?

In order to avoid opinion based answer I would thus specify: If you answer, please specify if what you say is based on your personal experience as a composer or on some other documents.

closed as primarily opinion-based by guidot, Tim, Richard, ttw, Todd Wilcox Apr 8 '17 at 6:23

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • Comments are not for extended discussion; this conversation has been [moved to chat] where it can hopefully be resolved. (chat.stackexchange.com/rooms/56850/…). – Doktor Mayhem Apr 10 '17 at 7:36
  • Vaaal, the reason this is on hold is because it is very much opinion based, and this view is supported by all the comments and discussion, which I have moved into chat. – Doktor Mayhem Apr 10 '17 at 7:43

I can't speak on behalf oa any other composers, though I think I know in many cases what they would say. Speaking only for myself, I think both things happen, at different times. I know there have been times when I've worked out a harmony that sounds right and later gone back to see what it "coincidentally" was. And there have certainly been times when I've thought, "an augmented would sound good there," and, yes, it did.

But names like "diminished" etc. are just NAMES. When you've heard enough and composed enough, you have a pretty good idea what something is going to sound like, whether you already know the name for it or not. Usually, you do. but not always. For example, I can never keep straight in my head which of the augmented 6th chords is associated with which nationality. But I know which one would sound good in a given circumstance. Later I might decide to look it up and see whether it's the Italian, German, or whatever. And I don't really care, or need to know, unless I'm telling someone else about it.

To put it all a different way, my inner "ear" is more important to me in composing than my knowledge of theory or nomenclature is.

  • 1
    Thank you, this is really close to what I was looking for. Of course I would like to know how "most" composers do, but your own personal experience is very valuable. – Vaaal Apr 7 '17 at 14:20
  • 2
    @Vaaal "most" is difficult; because processes vary between composers. Also different styles in general are more/less forgiving of a just "feel" based approach rather than a "thinking" one. You couldn't write "giant steps" (or even solo to it) without a very deep knowledge of explicit music theory. For other genres the reverse is true; "analysing" isn't good for that much more than describing what sounded good to you in the first place. Most music falls somewhere in the middle in that both are necessary. However, this may not be a "formal" knowledge of theory. Take Paul McCartney (cont.) – Some_Guy Apr 7 '17 at 16:11
  • 1
    @Vaaal By any standards the Beatles later output is some of the most complex and adventurous tonal music of its time (analyse penny lane or here there and everywhere sometime). McCartney has no formal music education, knows nothing about the "names" for devices, yet it's clear he would use "thought" in addition to "feeling" to write music like that. But if your past experience of experimentation, listening to others' music and songwriting tells you how a specific device will "feel" to you before you play it, where does that fall in an "intuition" vs "conscious thought" paradigm? – Some_Guy Apr 7 '17 at 16:17
  • @Some_Guy, that's a good observation. For answering your question, I would classify that as "unconscious" harmonic structure, EVEN THOUGH is "conscious" feeling of what's good or not. – Vaaal Apr 7 '17 at 16:26
  • 1
    @Vaaal- I think it's also important to remember that "traditional" theory wasn't developed until the early 19th century, so composers before that could NOT have thought in terns of dominant or diminished or any of those names for things. They knew what they were but didn't call them by the names we use. – L3B Apr 7 '17 at 17:02

Well, concerning Bach: he was an organ player (and concertmaster and director) in the baroque tradition, so he had to work a lot from scores written in figured bass where the player had to extemporize an accompaniment from the written-down harmonies.

That he was fussy enough as a composer himself to mostly write down everything explicitly rather than relying on the skills of the executioner does not exactly mean that he was unable to grasp the concepts of harmonization.

Similar things hold for pretty much all classic composers of renown: when they became famous due to being innovative, they weren't innovative instead of being skilled and learned but on top of it, and that was what made them a cut above the crop.

And that also goes for a lot of modern and experimental classics where people state "my cat could do that better", just like people think Picasso paintings could equally well have been done by a five-year old child and wonder whether Picasso was able to draw "properly" at all.

To break and extend the rules in a way that appeals to the learned, you have to master them. You cannot create in a vacuum: you are always in the context of tradition and convention, and to be relevant, you have to be competent with the context which the culture is based on. Transcending the rules in a purposeful manner requires being aware of them.

This does not mean that the composers were necessarily thinking in the exact terms and categories that we apply these days.

  • 5
    No offense, but this is totally off topic with my question, so I don't understand the number of upvotes. Maybe I was not clear. Quoting myself from another comment: when composing, would composer go like "uhm, here a diminished II would sound nice" or do they go like "uhm, THIS chord that I hear inside my head would sound nice... oh look, coincidentally, is a diminished II" – Vaaal Apr 7 '17 at 12:51
  • 4
    I'm pretty sure you meant "fussy", not "fuzzy". There is a difference. :) – Ilmari Karonen Apr 7 '17 at 17:21
  • @Vaaal working to understand why this has upvotes would help you understand how it answers your question. Just because you're not able to grasp the relevance doesn't mean it's not there. – Todd Wilcox Apr 8 '17 at 6:18
  • @ToddWilcox, I believe that comments with this tone (the one you are having) are very counterproductive. – Vaaal Apr 8 '17 at 9:04
  • Vaal - Todd is actually trying to help you with understanding what parts of your question might be answerable here, and how there are aspects of your question you may not have considered yet. Please take his comments that way. – Doktor Mayhem Apr 8 '17 at 13:26

Okay there is very little hard evidence for long dead composers but I think you are underestimating the musical abilities of people like Mozart and Bach. It is perfectly possible - and to my mind, likely - that Mozart envisioned complete pieces in his head and then simply wrote them down on paper. No trying things out to see what they sounded like or anything like that, he already knew.

So when he went, as a teenager I think, to the Sistine Chapel, and heard a performance of the Allegri Miserere (which was not available as a piece of music outside of the Vatican) he simply went home and wrote it all down, complete, after having heard it just once. Mendelssohn is supposed to have done the same thing some years later.

This ability to hear things internally obviously applies to Beethoven as well, which explains why he could still compose even after going deaf.

So I think that certainly the great mainstream classical composers all knew exactly what they were doing.

  • "Mozart envisioned complete pieces in his head and then simply wrote them down on paper." - You don't need to be a genius to do that. Even I can manage it! But in the context of the OP's question, the harmonic structure of music may not be very important - there are plenty of musical genres with no "harmony" at all, in the Western European sense of the word. – user19146 Apr 7 '17 at 12:21
  • This is, again, not what I was referring to. Mozart or alphzero or myself would be able to do that, even without a piano, WITHOUT having to be aware of the classical structure of the piece (or, at least, not all details of it)... – Vaaal Apr 7 '17 at 12:47
  • Let's say I hear two chords, I - V. I can either recognise the structural harmony and think "oh, they are I and V, and this sounds like Cmaj so here how you play it", or they could just feel it internally: "this two chords sounds like this and that...." and then only AFTER they feel it they could go "ah, these are actually I and V". However, I am not referring to feeling OTHER pieces structure, but their own, WHILE they are composing. – Vaaal Apr 7 '17 at 12:48
  • @Vaaal Mozart didn't write "classical music." He wrote contemporary music. And some of what Bach wrote was so avant-garde that it horrified his audiences, and his employers - that was one reason why he never landed a "top-grade musical job" in his entire life. A description like "classical structure" only has meaning with the benefit of hindsight - and arguably it doesn't really have any meaning at all. – user19146 Apr 7 '17 at 13:39
  • @alephzero remove classical from my previous comment. – Vaaal Apr 7 '17 at 14:09

One learns to compose by listening to a lot of music and then somehow taking the resulting music in one's head and writing it down / playing it / punching it into a computer etc.

If I listen to a lot of Beethoven in a row, my music will start sounding more like Beethoven, not because I am thinking logically, Beethoven would use a German 6th here, but just because he does it so it often, it becomes normal.

Also, you can learn how tension and release can be captured harmonically. But this is also based on experience and culture as what might be considered pleasant or tense in one cultural context might not resonate at all in another.

Some things that "sound bad" are related to consonance and dissonance. If you listen to enough Bach chorales you will recognize parallel 5ths or an unsupported 4th and say that "sounds bad". But then you can listen to enough pop music where it is normal and suddenly it sounds ok again.

Surely only composers like Schönberg actually composed by the numbers - or consciously avoided doing so (aleatoric composers). Everyone else just acquired a body of repertoire and built on that; the great ones succeeded in coming up with something equally good or better but the roots were there.

  • Atonal music gets a bad rap. There's a case to be made that it's actually one of the few genres where you specifically can't compose by the numbers. Even if you're not into atonal music, the crazy thing about it is that there is absolutely no harmonic structure to rely on, no "playbook"; there is no paint by numbers approach, and so you just have to rely on other elements. Somehow something like youtu.be/q7q9wtvMMgs doesn't sound "random" it somehow sounds like music. (serialism is no more a paint by numbers approach than a character restriction is a guide on how to write a good comment) – Some_Guy Apr 8 '17 at 0:34
  • @Some_Guy Berg Op. 1 isn't really atonal – heck, it even has B minor in its title! Sure it also has some weirder stuff in it, but if you want to talk “absolutely no harmonic structure” then this is not a suitable example. – leftaroundabout Apr 10 '17 at 22:24

There are a few different answers that I think are appropriate to the question you’re asking. The first distinction is between music that is composed melodically as opposed to harmonically. This contrapuntal style applies to all music composed before the Baroque period, as well as specific genres and styles that followed in the same tradition (most especially fugues and inventions). While these styles do have harmonic implications, such as tonic-dominant relationships, they are a by-product of following the common practice rules of counterpoint.

The next factor is one of historical context. Once chord theory, which was in itself just a codification of practices that emerged from the previous generation’s counterpoint rules, began to take hold during the Baroque period, composers began to explore “functional harmony;” this is the idea that there is a hierarchy of chords that operate in specific ways. You can hear composers experimenting with chord qualities and inversions throughout the Baroque period, and we can glean intentionality from the gradual exploration of these choices, not to mention the ubiquity of very specific cadential structures.

Your post mentions Mozart, for whom we have his letters and his collected manuscripts (hoarded meticulously by his wife) as proof of his process. Although he is iconic to the classical period of music history, during and after which comes music that I think provides the most interesting answer to your question, he was such a musical genius that his example provides both proof of intentionality in chord choice and possibly a refutation as well. Mozart began playing the piano when he was barely a toddler, and improvising short works in the contemporary style only a few years later (these were transcribed by his father and exist as his first compositions). This intense training from a young age, not uncommon for children that showed aptitude in music, coupled with the fact that music was the primary source of entertainment for members of the upper and slowly-emerging middle class, meant that the rules of counterpoint and theories of functional harmony were ingrained in Mozart’s psyche. Much in the same way that a professional athlete isn’t thinking about every single physical movement or learned technique while they perform spectacular feats, Mozart’s music (and that of many of his more esteemed contemporaries) was probably not hammered out chord by chord. The thing that makes Mozart an extraordinary example is that he could conceive large sections, sometimes entire movements, of a work with all their orchestral parts before writing down a single note (the research behind this is very convincing, discovered by musicologists scrutinizing types of manuscript paper, receipts, and timelines).

Beginning with Beethoven and essentially ending with Wagner, we see composers exploring the limits of tonality through pitch relationships. Outside of the development section of sonata form movements, acceptable harmonic areas up to this point in musical history included, in major keys, the tonic, dominant, and subdominant; and in minor keys, the relative major and dominant, although minor keys were more experimental compared to major keys and sometimes included others. Limiting the music to closely-related keys also limited the pallet of chords and pitches available to a composer. The intentionality of chord choice becomes clear in the way composers used logical but unconventional voice leading to reach distant keys. The modulations in late string quartet by Beethoven required meticulous attention to detail and obvious planning, also made clear by the musical sketchbooks that he maintained for posterity. Skipping forward, the musically infamous and functionally ambiguous “Tristan” chord from Wagner’s opera Tristan und Isolde shows intentionality in chord choice even before beginning to compose any significant portion of melodic material. The use of common tones in chords to achieve distant modulations, enharmonic respellings to facilitate unusual voice leading, and intentional disruption of typical harmonic motion (like moving backwards around the circle of fifths) all prove intentional design.

All of that to say, the answer is somewhere in the middle. It isn’t possible to precisely extrapolate someone’s though process, but the evidence proves that most composers are probably not thinking about chords on every beat of music, but almost inevitably doing so at important cadences, modulations, and other transitional or significant thematic moments.


You should probably be disappointed that no one seeing your question can speak on behalf of the sentiments of any deceased composer XD But I suppose they are completely aware of what they are doing. In the age of Rameau (about same as J.S. Bach), the harmony rules are already well known. Students often improvise when practising as a organist, and knowledge of composition was probably passed orally or at least through apprenticeship. By the end of 19th century, the existence of music school is already established, and the rest is history.

In fact, I suppose most competent composers improvise a little bit to think of a melody, and guess a suitable harmony, then examine the result using rules they know, and play on the piano to evaluate the effect, and feedback the whole process. If they internalize the process, the result is amazing.

  • I am also asking to the composers in this channel about their personal experience. Do they go like "uhm, here a diminished II would sound nice" or do they go like "uhm, THIS chord that I feel inside my head would sound nice... oh look, coincidentally, is a diminished II" – Vaaal Apr 7 '17 at 12:49
  • 1
    This is a bit misleading. Of course there were "harmony rules," but Rameau was the person who invented the concept of inversions of chords, and quite likely J S Bach never read his book where he published that idea (certainly his son C P E Bach completely ignored Rameau's ideas in his writing on music theory). Analysing Bach using terms like "a diminished II chord" is a total anachronism, compared with how Bach thought about what he was writing. – user19146 Apr 7 '17 at 13:44
  • diminished II was just an example, I don't know if THIS PRECISE concept was known to Bach, but he probably knew about tonal relationship, right? – Vaaal Apr 7 '17 at 14:06
When composing music, does a composer go like: "Uhm, here I should put the V chord in second inversion as is the thing that follows best the II chord [I AM MAKING THIS UP]" or like "ah! This sounds very good in this position" [and only AFTER THAT maybe analysing from a harmonic point of view]. Which one would they MOSTLY use?

Does a carpenter go like "Uhm, here I should be using a size 4 screw driver as it is the thing that best fits this screw head" or like "Ah! This tools fits very good in this position" [and only AFTER THAT maybe analysing what he has actually taken into his hand]?

Working with the tools of a craftman's trade competently will not take the attention span your first alternative suggests, nor will the composer be surprised at what he did like your second alternative suggests.

A writer does not spend a significant amount of his attention on being grammatically correct. That's part of the basics enabling him to be a writer in the first place.

  • is this part of your personal experience? Or this is what you think is going on in a composer's mind? – Vaaal Apr 7 '17 at 14:39

Not the answer you're looking for? Browse other questions tagged or ask your own question.