I've been told by many people that music theory is not normative; it doesn't tell you what to write or do in music to create something beautiful that sounds good. I've been told that it's strictly descriptive; that is, it describes a set of observed patterns in music that (someone) thinks sounds good, or that is frequently written.

So with that in mind, I want to ask about theory and composition. How is theory supposed to inform composition? Some musicians seem to have a deep theoretical knowledge, and regardless of what they would say their mental process is, this seems to affect the way that they compose, since they can't "un-learn" certain patterns. On the other hand, there are untutored musicians who know nothing about theory, and who in some sense are more "free to break the rules", being unaware of what the rules are.

EDIT - This question has already gotten a lot of attention, and several good answers. I'm starting to think the original question is a bit flawed; here are three big things I've gleaned:

  • Several answers point out that people who don't know any formal theory still adhere to many patterns described by formal theory, in the way that a child who speaks English knows sentence grammar without knowing the terms of academic structure of grammar. In this way perhaps, they're "using theory" without having a conscious awareness of rules and terminology like chromatic mediants, modulation, etc.
  • Another perspective here is that focusing on the connection between theory and composition is the wrong focus, because what matters is whether the result is good, pleasing, unexpected (pick your descriptive term). If it is, then it doesn't matter what the theory connection was. If it isn't, then clearly it also doesn't matter what the theory connection was.
  • There's a fair amount of concern about the interplay between theory and creativity. Does knowing theory make you more creative, or less creative? There's no consensus, it appears the answer is "both, depending on who you talk to".
  • 7
    I used to think that, in C keys, a D chord before a G chord sounded absolutely amazing, but I had no idea why. Additionally, I wasn't even sure whether anyone agreed with me on this. Then I learned about the theory behind secondary dominants, which taught me why this sounds good, and how I can reach a similar sound in many other ways. I never thought music theory was very useful until that moment. – Lee White Jan 20 '15 at 18:29
  • 1
    Your first paragraph is true, except in cases of modern protestant liturgical hymns... once you know a few basic chord progressions, you will know every Christian worship hymn ever :-D – TylerH Jan 20 '15 at 20:00
  • 1
    Only if you know the rules, you can conciously break them. – leonbloy Jan 20 '15 at 21:05
  • Music theory prevents you from re-inventing the wheel, which is a time-consuming process which will be recognized by your peers as clichés. – Sanchises Jan 21 '15 at 20:39
  • ""using theory" without having a conscious awareness of rules" -- this is a common guess. But conforming to a descriptive theory isn't necessarily using it, any more than a rock is "using the theory of gravity" to fall ;-) In such cases the theory is an approximate representation of what happens, it isn't what happens. – Steve Jessop Jan 22 '15 at 11:48

12 Answers 12


On a practical level, knowing some theory can be useful to a composer in certain circumstances.

First is when a composition is not as interesting as the composer would like it to be. He/she would like to evoke something different than whatever the music inspires at that stage of composition. Examining the melody for how it conforms to standard scales, examining the harmony to detect whether or not it is feeling hidebound to predictable patterns and progressions, then contemplating how to mix it up with shifts in key feel, introducing certain accidentals - all this can be useful, and I can attest to it from personal experience.

Second is when one has composed a passage, and the harmony is not obvious. Analyzing the relationship between a melodic passage and all the possible ways it can be harmonized (Does it lie in a standard triad? Could it be an accidental?) can be interestingly informed by theoretical knowledge.

I like to think of theory and composition having a similar relationship to that between the composition of poetry on one hand and rhyme, meter or other craft-centered knowledge on the other: It can be a limitation, sending the composer down a well worn channel to a banal destination, or it can be a spur to creativity that lifts one out of a rut.

Of course, it can never substitute for the inventiveness that composition demands of its practitioners. That is why composers of tremendous accomplishment have done without such knowledge, and yet excelled.

| improve this answer | |

It's rather like language. Treating the rules as prescriptive allows you to always be generally understood (or compose something non-irritating). You need to understand the rules in order to know when it is OK to break them, even though you could accidentally break them and still make a comprehensible sentence (or pleasing melody) without knowing why.

Because the rules are in fact descriptive and often involve natural laws (i.e., the math behind the music), music can come quite naturally to people. Like the other answers say, composers who have little knowledge of theory are still unknowingly following many rules. (Or, more accurately, the rules describe them already.)

Whether you arrive at your composition "naturally" or from some laid-out process doesn't really matter if you/others like it. (Of course, it's totally valid to purposefully compose unpleasant works as well.) Both are effective methods for different people, but neither stands alone. Following the rules only works in combination with some sort of melody or desired sound already, since otherwise you'd just churn out the same thing every time. Conversely, you can't compose without any rules because many of them are already built into your brain, and our brains are what defines "music" in the first place.

| improve this answer | |
  • I appreciate your response, and I'm really not trying to give you a hard time; but the question was "How is theory supposed to inform composition?" - your points are well-taken, maybe the process doesn't matter as long as the result is good. So are you basically saying that there isn't a connection between the two, or that if there is, it doesn't matter "as long as the result is good"? – FrobberOfBits Jan 20 '15 at 20:41
  • @FrobberOfBits Maybe I am not totally sure what you mean by "supposed" -- "intended" or "able". Whether theory should inform composition is up to you, and what I'm trying to get at in my answer. If you meant how can theory inform composition, I would recommend memphisslim's answer; at this point I don't feel comfortable going back and adding what he said to my own answer. – user28 Jan 20 '15 at 22:01

First off, the notion that you can write more freely if you "don't know the rules" is an unfortunate fallacy. When I hear guitar players saying that they eschew learning theory or how to read sheet music because it will "stifle their creativity" I think, "lads, you're trying to run a marathon with a boat anchor strapped to your ankle."

Functional harmony (western music theory) was constructed empirically, by recognizing the patterns that everyone gravitated towards and capturing those in a descriptive semantics.

When you compose having a background in music theory, you possess the tools to compose "with intent." The theory essentially describes patterns of tension and release, and patterns of consonance/dissonance. Music theory is not there to make it "sound nice," it's there to provide you with a tool to achieve specific intent.

All that said, it's obvious that there are many songwriters who have no formal training to speak of who come up with wonderful music. The reason they make wonderful music comes out of their wonderful mind, not out of their ignorance. They actually DO use theory, but it's born of their own sensibilities and listening.

| improve this answer | |

Theory is both descriptive, and, to the extent that it gives an accurate description, also a bit prescriptive (but only to that extent). It might be correct to say that theory doesn't provide rules so much as rules-of-thumb. What theory does is provide a set of tested solutions to common problems. If, in a piece written in common practice tonality, you find yourself at V/vi needing to get back to I, chances are good that your theory courses have taught you some very serviceable progressions to use to do so. If, in a twelve-tone piece, you need to accompany some voices written using one row form with voices using another row form without incurring any unisons or octaves, well, you look up combinatorial rows - the matter is quite simple.

The problem with trying to use theory for much more than that lies in two spots:

  • Theory covers commonality, what happens the majority of times in a given situation, which is fine if you aren't attempting much more than Gebrauchsmusik or musique d'ameublement. If you like setting challenges for yourself, you'll frequently find yourself in spots where the "rules" no longer apply nearly so well. In fact, most of the rules came about when someone had to write something extraordinary, something that was outside of common practice, in such a way as to make it intelligible to the audiences of the day - in other words, to make such a departure from the rules compatible with the spirit of the existing musical language, to make such departures normative. (This is a good description of so much of Haydn's output.)
  • Because of that, theory is reductive. Because of the emphasis on commonality, theory is sometimes obliged to ignore a lot of what particularises a specific well-made work. Sonata form is a case in point: sonata was not so much a form as a way of working in a language that expected a move to a dominant region, and a resolution of that move, as part of its normal syntax. Yes, you can often find first and (sometimes "feminine") second themes, transitions, full recapitulations, discursive "development sections", but it was only when the language changed so that dominant polarisation lost its force that the "form" ossified into what we find in textbooks now. (See Charles Rosen, "The Classical Style", "The Romantic Generation" and "Sonata Forms". Rosen's writing on the subject is marvellous.)

Some musicians seem to have a deep theoretical knowledge, and regardless of what they would say their mental process is, this seems to affect the way that they compose, since they can't "un-learn" certain patterns. On the other hand, there are untutored musicians who know nothing about theory, and who in some sense are more "free to break the rules", being unaware of what the rules are.

The former is true to the extent that these musicians see theory as a panacea, a kind of "silver bullet" to slay the complexity encountered when composing music. If you are interested in drawing out the implications of your musical ideas, and doing so in a language that is intelligible, you will frequently find yourself feeling your way forward, and these all-encompassing theoretical frameworks become, at best, guidance, and, at worst, detriments. Personally, I find that the older I get, the less I know for sure, the more contigent my ideas regarding theory become.

As for untutored musicians, they frequently succeed to the extent that they have internalised the prevailing musical language, since they don't live in a musical vacuum. However, when you have an autodidact like Mussorgsky, who was attempting a synthesis of classical and Russian folk traditions while eschewing "book learning", you can sometimes find such a person having occasional problems adequately conveying his ideas (and, in Mussorgsky's case, it led to a lot of people, rightly or wrongly, feeling a strong need to edit his work). A knowledge of theory sometimes allows you to find solutions to the specific problems immediately at hand by analogy to existing solutions to common problems.

| improve this answer | |
  • Thanks! This is a great answer. I think what I'm coming around to is that my question is based on a set of assumptions about what theory is and what its relationship to good music is, that I need to revise. I appreciate you taking the time to write this. – FrobberOfBits Jan 21 '15 at 13:06
  • @Patrx2 I agree - very well put. But don't you think prescriptive is the opposite of descriptive. I did not quite understand what you meant to say in your first sentence. – Rockin Cowboy Jan 21 '15 at 20:29
  • 1
    @Rockin Cowboy, they're not completely mutually exclusive. Let me put it another way: theory is useful for telling you how something can be done to the precise extent that it is successful at describing how something has been done in the past. – user16935 Jan 21 '15 at 21:44

In math, you are supposed to know 1+1 = 2. You do not need to know math in a formal way to know when you have two apples. As math becomes more technical, common knowledge is less equivalent, even while it is still relevant (What is the thrust required to leave the surface of the planet, or what is the compound interest of your account over the next 6 years?)

I know music is not math, but it is similar in this way: You do not need to know music theory to write a simple piece of music that follows the rules. You do need to know the rules in order to achieve a complex musical work that connects to the common musical grammar. The good news is that simple music can often sound really good. That does not negate the riches to be found in deeper study of music theory.

Some can also internalize many of the rules of music in common use, by listening. This is equivalent to grammar. A person can follow the rules of common use in grammar without ever going to an English class, and without even thinking about it consciously. This does not mean that attending an English class is useless, what it does mean is that it is possible for some to follow the rules without it.

Understanding music theory gives you the opportunity to discuss the rules in a forum like this, and study and research them further and find answers to questions. It gives you a larger deeper world to draw from and a view to standards that have already been set.

Its harder to break the rules if you do not know what they are. When I was young I wrote pieces that were meant to break the rules, but since I did not know the rules, no rules that I meant to break were actually broken.

I don't think you would learn unbreakable patterns that stick with you forever, unless it turns out that you like said patterns and want to continue with them on some level. Win : Win.

| improve this answer | |

You can use theory to form a framework which underpins your composition. The framework provides a basic structure that guides the composition, but the details of the composition distinguish it from other compositions, even those that use the same framework.

| improve this answer | |

How is theory supposed to inform composition?


On the other hand, there are untutored musicians who know nothing about theory, and who in some sense are more "free to break the rules", being unaware of what the rules are.

I contend the fact that good musicians break the rules. If you break the rules for the mere sake of breaking them, it will definitely not lead to good music. Even impressionists who broke a whole lot of conventions did not do it for the mere sake of doing it, but had definite stylistic ideas they wanted to follow.

There may be musicians who have no training in theory, but still apply the things theory teaches to their music (even if they are unaware of doing so). James Hetfield, for instance, if he would ever feel the need to do theory, would instantly know the proper use of rhythm in melodies. He is a great rhythm player and uses the principles of good rhythm in his music regardless of whether his education is formal or not.

Even if you do think of impressionists and how they made their music, there was not a great room for this type of music to evolve and grow. That was the great deficiency of the genre; it lacked the manner in which other musical eras / genres evolved, and this had a great deal to do with the way in which it discarded long held conventions.

| improve this answer | |
  • 2
    Can you elaborate on "completely"? Below you gave examples of people who weren't trained in theory but were using it anyway. This wouldn't be theory informing composition, it would be theory fit to a composition after the fact. For people who do know theory, exactly how do they use it to inform their composition? I'll accept "completely" as an answer, but it needs elaboration. – FrobberOfBits Jan 20 '15 at 15:42

As other's have suggested, it is not necessary to know any theory at all to create great musical compositions. Music follows some basic laws of nature that are innate not only to humans, but some studies have even suggested plants can respond in predictable ways to music. Theory is merely an attempt to logically explain why some music sounds good and some notes sound good together, while others do not.

That being said, theory can indeed be used effectively to, as you say, inform composition, should one have a desire to use it in this way. For example, theory tells us which notes blend well together to make music that is pleasing to the human ear. So if we are composing a piece in C Major, we know from theory, which notes are available to us if we want to remain within the safe harbor of tried and tested theory about which notes blend well together in that particular key. Knowing which notes will work in our melody also tells us without trial and error, which chords can be formed from the notes in our chosen key.

We can also apply theory to better evoke the desired emotional response to our composition. Through theory, we understand that minor keys seem to evoke a more melancholy or sad feeling whereas major keys evoke a happier feeling.

If our goal is to compose a piece of music that will involve multiple instruments played by different musicians, then sticking with tried and trusted theory, will make it easier for the other musicians to either learn their parts, or instinctively follow the pattern because it fits within the framework of music theory, which is in reality, merely a descriptor of why a musician would naturally gravitate towards a certain note or chord or rhythm.

Allowing theory to provide the framework that defines certain parameters of the composition will help insure that the musicians who must perform their part of the composition will not get off track by following their natural instincts. It also gives you a degree of certainty that it will sound "right" and be satisfying to the listener.

In summary, if we wish to assert that music theory IS supposed to inform composition, it does so by defining the key, the tempo, the rhythm, the notes available, the chords available, and other parameters that will best convey the intended feeling or emotion to the listener in a pleasing manner - while being easily interpreted and performed by the musicians or instruments that the composer wishes to use to recreate the musical composition.

| improve this answer | |

You need to know theory because:

  • Music composition is a craft. You may very well do something intuitively, but that doesn't change the fact that you need to learn the craft. More often people have an intuition for melody and harmony but they rarely have intuition for form, and form essentially is composition. Without form your music is at best sound design and at worst a blob of incoherent sounds. Learning theory can make your music more coherent and listenable. It might not, but it can.
  • Music theory gives you awareness of what you are doing, so that you don't naively think that you've just done something no one else has done before.

Being unaware of the rules doesn't automatically give you an advantage, knowing theory doesn't either. What gives you an advantage is hard work and some talent.

| improve this answer | |

Everyone has is own music theory, often one for each music style, and music theory can sometimes be thought as a generative language.

Ideally, the music theory should predict what can come next after, for example, a given succession of chords : letting you choose among the 10 possibilities permitted by the music theory, instead of searching for the next chord among an infinity of possibilities.

There is something magic about music, particularly the melodies. But that magic part of music (that is related to imagination, the concept of beauty, human capability to dream..) represents only 40% of the whole, the rest being algorithmic. That algorithmic part, related to grammatical rules, is exactly what music theory is about.

In my opinion, missing one of these two parts of music : magic + algorithmic will make you incapable of composing music efficiently.

| improve this answer | |
  • Bach, Mozart, Beethoven, Chopin, they all have an obvious algorithmic component in there music. And today, electronic music probably represents the best example of music with a huge algorithmic part. – reuns Sep 2 '15 at 22:20

I am going to say some heretic thoughts and think that have the right to do that as I already have presented working radical new music system which is quite works from the very beginning and needn't contemporary music theory. ( http://home.arcor.de/yuri.vilenkin/new-music-atoms.html http://www.youngcomposers.com/forum/t33636/the-new-structure-of-muscal-note/ http://forum.emusictheory.com/read.php?5,16533 ).

Of course it doesn't mean that possible alternatives make unimportant the knowledge of established music theory as it is necessary for learning and understanding of already created music practice.
Nevertheless it is worthwhile to limit oneself to several most important authoritative works as, for example, not ago anew disclosed "Guide to the practical harmony" of P.Tchaikovsky, who isn't occupied with inventions of vague notions as "tonality" and doubtful speculations. Particularly he wrote:

" It is important that by thoroughgoing explanations of rudiments we should dispel at the start the superstitious awe that prevails regarding so-called 'Theory' or 'Thorough Bass'."

Namely so called " theory". The great deficiency of established music theory is that it presents itself as bearer of eternal truths. In reality the its most important part the harmony theory is about chords and progressions, which appeared by spreading of key instruments and may become unnecessary by further technical and scientific progress.

Now computer technology and results about of human music perceptions obtained by physiology and psychology permit to build music system which is directly oriented on human sensations of scientifically verified music components (concerned to harmony), that is; of melody, dissonances, fundamental frequency, smoothness.

| improve this answer | |

Theory informs the craft of composition. Limitations in the development of craft can restrict the musician’s expressive potential within the art of composition.

Composition as Craft: virtuosity developed via music theory as generalizations, descriptive rules and patterns of style, technique, and method; developing understanding and excellence in the practice (i.e. the practice of music, law, medicine).

Composition as Art: a practice/process existing beyond theory, measurement and rules: imagination, intuition, creativity, exploration, curiosity, aesthetics; awareness of meaning

Expressive potential in the art of composition and performance is limited by:

  1. Expertise in craft: how, what;

technique, theory, practical experience historical awareness of styles, methods, techniques/technology, analysis, imitation, execution, proficiency, virtuosity, technique development, intellectual development; didactic and experiential learning; development of expressive means (toolkit).

  1. Artistic development: why;

life experience, emotional, perceptual and aesthetic awareness, intuition, imagination, insight, curiosity, innovation; a search for meaning and aesthetic expression beyond current limits.

  1. Intrinsic/Extrinsic factors: talent, intelligence, creativity, motivation, persistence, opportunity, support, etc.

AI (artificial intelligence) can generate music compositions reliant on rule-based styles, techniques and theoretical methods of composition (i.e. aleatoric, stochastic, algorithmic, serialisms, fugues). Increasingly, the rule bound aspects of the craft of composition (styles, forms, techniques, etc.) are being replicated and simulated by technology.


In a sense, AI technology demonstrates the generative possibilities of composition as craft only. That is, based on theoretical models. And this illuminates to a further degree many of the differences existing between the craft and art of music composition.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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