Comment by danShumway | Hacker News Reader

Comment by danShumway | Hacker News Reader

Parent Full thread

It's technically correct while also being practically useless.

There are no fundamental rules to grammar, but we still teach writers what a simile is. Even in spelling, English has no consistency about a lot of this stuff. Are spelling rules useful though for non-native learners? Heck yes. Do we still have grammar books? Yes. It turns out that learning the exceptions is sometimes faster than memorizing everything. Music is a communicative medium based in large part around social norms of what combinations of sounds tend to feel good to people who grew up in a specific environment, and music theory is an attempt to break apart that social consensus into basic rules that can be imitated and built up into more complicated genres.

The rules are somewhat arbitrary and made up, but that doesn't mean they're not real, and that doesn't mean you can't sort-of derive some of them from more basic rules or from basic principles.

Music could have gone in multiple directions, but for whatever reason Western music went in the direction it did, and many people who are trying to learn music aren't trying to learn the full spectrum of every direction music could have gone. They're trying to learn how to imitate Western sounds, and for that purpose specifically music has rules -- many of them derivable from other rules, flexible and breakable as they may occasionally be.

They're not rules unless your teacher is an asshole.

Music theory is a language used to be able to have conversations about harmonic and melodic aspects of a piece of music. Rythmic aspects are less well supported (that's putting it mildly) by western musical theory, but the same is true there for whatever formalism or terminology/nomenclature you might pick: it's not a set of rules, it's a language to facilitate non-musical discusison.

I think this is a distinction that only matters if your definition of rules is such that they can never be broken or played with. That's not typically the way I think of rules, and I don't think I'm alone in that. Even programmers understand that a lot of programming "rules" are actually just heuristics about how to produce good code.

As another example, I've spent a ton of time learning game design rules, and all of them are optional, but they're still useful. "Rules" in this context means, yes, communicative terminology, but communicative terminology about particularly effective ways to build commonly understood musical motifs/phrases that affect Western listeners in somewhat predictable ways.

Music theory is a language for talking about a language, music itself. It is partially social convention that leads us to have a 12 tone scale, and it is definitely social convention that leads us to call that a "scale". However, the average Western listener will respond to the notes of that scale in predictable ways, and there are "rules" that you can learn that will allow you to more easily and predictably manipulate that listener's emotions and communicate broader ideas through your music -- many of those underlying rules about tuning, ratios between notes, and so on are derivable from mathematical principles or at least describable in mathematical terms, even if ultimately the reason why listeners respond to some of those ratios is social and arbitrary.

Of course, the rules are not concrete or immutable, they can be broken and often are. But the majority of rules we learn in most subjects are not concrete.

Similarly, there are no concrete rules in writing, and the terminology we use to describe story structure is arbitrary and made up. However, learning the "rules" of writing will make you a better writer for typical audiences that live near you, and those rules are expressed through common set of terms and concepts that many professional/hobby writers have decided to use -- many of which can be partially derived or explained by talking about psychology or history or whatever.

Sure, these are not immutable, scientific principles baked into the heart of the universe, but:

A) the author goes out of her way to say that she isn't claiming that, and

B) breaking down rules and building them back up from different starting points or looking at them mathematically is still a reasonable thing to do with rules that have a social origin, and

C) even though a lot of why certain chords sound good is baked into culture rather than biology, it's also a kind of strong claim to say that all of it is purely social. But I don't think it would matter much even if it was purely social, people who do music theory for a living still talk about math sometimes.

Well, I wasn't really thinking of much in TFA as "the rules of western music theory" - it didn't really get that far. There was another comment that talked about how TFA didn't cover things like chord voicing and inversion, and there is just so much more that wasn't covered that really forms the meat of "the rules of western music theory". TFA really just covered "one basis for 12T, what scales and chords are", which is barely anything to work with.

"Western musical theory" is full of ideas about harmony (OMFG, so many rules), often with non-musical semantics overlaid on top of the actual musical elements. These are the "rules" that you get play with as a musician, and the very best of them do in fact break them frequently (but expertly). Why do so many people remember "Take 5" ... because it's in 5 not 4! Why do people consider Coltrane to be a genius ... because of the games he played with harmonic relationships mostly connected to the circle of fifths but deeply subverted. Why do some of us still celebrate the "genius" of Schoenberg, Stravinsky or Bartok ... they upended traditional rules about harmonic resolution, even the very notion of tonal harmony in some cases. Even within this thread, we see people noting a striking detail of a recent Adele song that consists of (almost certainly deliberately) singing somewhat off-key to strong effect.

There is a huge amount of recorded music that plays entirely by "the rules" (the ones that go way beyond TFA), but a lot of what people think of as musical genius is precisely the stuff that flouts the rules (with enough knowledge of the rules to make this work).

> There was another comment that talked about how TFA didn't cover things like chord voicing and inversion, and there is just so much more that wasn't covered that really forms the meat of "the rules of western music theory". TFA really just covered "one basis for 12T, what scales and chords are", which is barely anything to work with.

I've commented to the same effect elsewhere, but people really underestimate how much the "barely anything" notation concepts are a real barrier to people who are unfamiliar with the domain. https://xkcd.com/2501/ comes to mind here; TFA is 5000 words and ends with multiple links to further tutorials and reading. That's a completely fine place to start. If you want to learn the rules of western music theory and get into the meat of what you're talking about, it is going to be a lot more work if you don't know what a scale or a chord is, and I don't see anything wrong with teaching those basic concepts from a mathematical perspective.

----

> with enough knowledge of the rules to make this work

That's the part that the author's target demographic wants. They want to be able to either:

A) imitate the rules well enough to write something passable (say for game soundtracks), but aren't looking to innovate, or to

B) learn the rules well enough to innovate.

In both of those cases, even the most basic notation concepts like "what scales and chords are" is a serious barrier to entry for many people.

> I don't see anything wrong with teaching those basic concepts from a mathematical perspective.

One thing that often goes "wrong" with this is that the mathematics of frequency ratios is a huge barrier that's most often irrelevant to actual music making. If you literally know nothing about music, there's a good case for just letting the tuners deal with it for the time being, and starting from the old Do, Re, Mi etc. that teaches you both solfège (sight reading/aural skills) and the musical syntax of scale degrees. Then sing a whole lot of music in (movable do; fixed do is pointless except for specialists) solfège and try to make up simple embellishments and variations on what you sing. There's a "programmer's favorite" way of learning these too (these are called diminutions) based on the scale-degree leap that you're traversing, hypothetical elementary operations of musical syntax and whatnot; but good musical intuition will always be helpful. Guess what, now you're well on your way to improvising simple music at the keyboard, and later on you can even get started on learning counterpoint without being lost in all the details. Because, unlike actual college students who are forced to take a counterpoint class as part of studying bookish "music theory", you'll have the fundamentals down pat.

If that works for you, great. But I don't think the article is implying that learning music has to start with teaching people about frequencies, it's just saying that some people (like the author) have found it helpful to latch onto.

> If you want to learn the rules of western music theory and get into the meat of what you're talking about, it is going to be a lot more work if you don't know what a scale or a chord is, and I don't see anything wrong with teaching those basic concepts from a mathematical perspective.

The problem with teaching them from the particular mathematical perspective taken in this work is that...it doesn't actually teach them, and it throws up it's hands and says I don't really know about fairly basic stuff. This isn't an alternate pedagogical route chosen by someone who has a different view of how to get people up to speed for the domain, it's a smattering of trivia that isn't directed at learning the rest because the author doesn't understand the basics, much less have a particular pedagogical approach to them.

There's a bit of jumping around here, because I'm responding to people who are telling me that mathematical models for music shouldn't be taught at all and that we shouldn't use the word "rule" in music theory.

I think that's a separate conversation from whether this article specifically should be the entire basis for someone learning how to compose music. I agree that I would not point someone at this article and say, "this will be enough to get you on the road to learning how to compose music", I would want something more involved by someone who has more experience.

But that's different from saying it's wrong for the author to talk about sound frequencies.

----

To your point in specific, I don't see where the author ever claims that this is an alternate pedagogical route to learning music, the author actively discourages people who already know music from reading the post. The post explains a few basic concepts like what a scale is -- and fairly accurately (or at least about as accurate as most of the other explanations that you'll find online). It openly works through the stuff that the author understands, and openly admits what the author doesn't understand, while sympathizing with the target reader that it's hard to pick up a new subject when it sounds like everyone else is speaking a different language from you.

So basically, it is every single technical blog post written on any subject by anyone who is openly learning about a thing online, and that is something that should be encouraged, not derided.

I also still kind of disagree with people who are saying that this is just trivia or too basic to even talk about, commenters are still underestimating how little normal people know about music.

If people come out of this article understanding stuff like:

- There are 12 "steps" in an octave

- An octave is doubling the frequency of a pitch

- Written down on a staff, we compress those 12 "steps" into roughly 7 spaces.

- A scale is 7 different notes.

- Because of weird ratio stuff and social consensus about which "steps" in an octave are most commonly used, some parts of the scale move 2 "steps" and some move 1

- Transposition exists as a concept, you can play the same song starting at different pitches

- Notes like B# and C can overlap, except some really professional players might treat them differently because it turns out the math we use for different pitches doesn't completely work out correctly in all scenarios.

That's all stuff that people who are unfamiliar with music don't know. Okay, the author doesn't really understand what's going on with minor keys, but this isn't a textbook, and there is value even in something as simple as "an octave is doubling the frequency of a pitch."

I'm weirded out by how upset HN is being about an amateur blog post with reasonably correct information by someone who is actively trying to teach themselves how to write music. This is exactly the content that we should want people to write about online, and it's written in exactly the style that we regularly encourage bloggers to write in when they're exploring new concepts/domains.

I remember Take Five mainly for the bass in the background during the drum break. But then, I like listening to rock music in Just Intonation, so I'm not typical.

Wholeheartedly agreed w/@danShumway - wish I read this before typing my own response. I might have saved myself some time :)

That's not entirely true. In fact, the Sound-Harmony-Melody-Rhythm-Growth model for analysis explicitly calls it out as a critical building block of the music that we hear and experience. Like any subject, it takes some study to get to the interesting bits.

For example, the vast majority of people who have taken Music Theory 101 have completed coursework that is the equivalent of "Hello, World!". That is barely scratching the surface and certainly give the aspiring programmer the skills necessary to create something non-trivial. The same is true of music theory. You don't instill in someone the musical understanding of Brahms with one music theory class.

I think that there are rules, but they're less about a teacher being an asshole. Musical styles evolve over time. We wouldn't have classical without baroque, and we wouldn't have romantic without classical. The style itself imposes a "rule" that is really a best practice. How do we in technology treat best practices? Well, they're basically rules that we shouldn't violate without careful thought. The same is true of music. People who study basic music theory learn about the cadences used most commonly in church music ... V-I ... IV-I ... and the deceptive I-IV-V-vi cadence! You feel where music is going because you've learned the rules by listening to other music and embedding yourself within its best practices. V-vi is deceptive because it violates a best practice, but used effectively, it works.

As music evolved, and as the world became smaller in terms of ease of travel and exposure to other cultures' music, we started to see new influences. Debussy and others were influenced by indonesian gamelan music. Without that influence, we might not have the jazz that we have today. Why? The impressionist style that Debussy practiced created floating pillars of sound, chromaticism, etc. that became important elements of early Jazz.

It's all connected and has evolved over time, just like technology. Personally, I do think they are rules, but like most rules, they're meant to be broken.

There's a world where no one thought to put the Amen break in a sampler, and that's a much less interesting world.

When giving this type of explanation, we want to emphasize that the rules are arbitrary and made up because the dominant way of teaching music theory from the 1920s through to the last decade or two has been to claim it as scientifically or mathematically "optimal" in some way. It isn't.

Starting off an explanation of music theory by talking about whole number frequency ratios is like starting off English class with a diagram of the glottis.

Do you feel like the article isn't doing a good enough job getting that lesson across?

> Western music has twelve distinct pitches. This is somewhat arbitrary — twelve has a few nice mathematical properties, but it’s not absolutely necessary. You could create your own set of notes with eleven pitches, or seventeen, or a hundred, or five. There are forms of music elsewhere in the world that do just that.

[...]

> I get the feeling that treating the whole chord/key ecosystem as a set of rules is like studying Renaissance paintings and deciding that’s how art is. It’s not. Do what you want, if it sounds good. I’m gonna go try that. Consensus seems to be that the real heart of music is managing contrast — like every other form of art.

----

> Starting off an explanation of music theory by talking about whole number frequency ratios is like starting off English class with a diagram of the glottis.

I would say it's more like starting out an explanation of color theory by talking about wavelengths, even though you don't really need to know any of that to understand composition, and even though most of the rules of image composition are contextual to specific cultures and aren't hard requirements that artists have to obey.

The math behind color is definitely not a requirement that artists should be forced to learn, but some people find it helpful to know what mathematical formulas were used to create a given color wheel.

Reply on news.ycombinator.com