Hacker News Reader

Comment by dahart | Hacker News Reader

> I always wondered why musicians keep up with the conventional musical notation system, and haven't come up with something better (maybe a job for a HNer?).

You're not alone, this is a common reaction to music notation by engineers; a lot of people have wondered the same thing, even here on HN. For example https://news.ycombinator.com/item?id=12528144 https://news.ycombinator.com/item?id=12085844

I see some great responses, but I wanted to add that you have to keep in mind that tons of people have actually tried to make a better system, and nobody has succeeded. That should give you enough pause to ask why and consider the possibility that the system we have is really good in a way that you haven't recognized yet.

I think the problem is that difficult to learn and bad are easily confused. It is difficult to learn.

Also keep in mind that music notation has undergone many iterations, and it represents developments over hundreds and hundreds of years and covers every instrument under the sun - the breadth of what it has done throughout history and what can do might be hard to see.

View on HN

>I see some great responses, but I wanted to add that you have to keep in mind that tons of people have actually tried to make a better system, and nobody has succeeded. That should give you enough pause to ask why and consider the possibility that the system we have is really good in a way that you haven't recognized yet.

I think that this is the incorrect way of looking at it. I suspect it is less that the traditional notation system is highly evolved and effective, and more that getting a critical mass of musicians to transition/relearn/teach/translate into a newer system is incredibly difficult.

For instance, while Imperial units aren't without some advantage, they are pretty generally inferior to the Metric system. But the US hasn't really switched because it requires a significant level of coordination and control that simply isn't easy to access. And getting musicians to learn and teach a brand new, objectively better system would be much much harder.

Music notation is to music as qwerty is to keyboards?

No, qwerty for keyboards is more like the lay out of the 12 notes on an instrument, and there are many instruments.

Music notation is more like a programming language. The score is like a program that you can read/interpret and play.

You have a great point, that getting the world to switch would be very very hard. But it's not black and white, you can't compare that to anything.

If there's a viable alternative to music notation that you know of and is superior to what we know as standard western notation, feel free to share.

Your choice of example is interesting, considering metric has won, and the US is switching slowly.

But there is no incorrect way of looking at it, music is an art. Standard notation is highly evolved and effective, it has been iterated on for millennia. Getting a critical mass of musicians to learn a newer system would be incredibly difficult. Both are true, and you can't compare them and say that one is "more", that's flatly not true in any meaningful sense.

I hope people don't think I'm being brusque here, but these comments are a classic case of an outsider looking at the system, admitting to be lazy and wondering why the rest of the world differs from their expectations vs. asking musicians what they think.

At its core, musical notation is succinct: a mixture of logic and unique symbols. Note markers are isomorphic to pitch. Rhythms subdivide with vertical lines. Special symbols and brief phrases denote beginnings, ends and loop points. (They're not usually in English) Geometric figures indicate volume and speed changes.

A competing system in my purview is "tracker" notation. It's vertical and generally only used on machines, but hand writable: It looks like: C-3 Eb3 G-3 Bb3

I have the same feeling. Music notation might be hard to interpret sometimes, but none of the alternatives actually solve anything. They do however introduce a whole lot of questions.

I think a valid comparison is the regular alphabet. It is, after all, a coding system for language in the same way that notation is a coding system for music. Most of the problems of that coding system (my pet peeve is english spelling) generally stem from conventions rather than problems with the alphabet (italian and german is much easier to spell correctly).

There might be some interesting alternatives (hangul!), but those systems come with their own share of problems and generally have no big benefits. I actually believe that musical notation is better fit for it's task than our current coding system for language.

> the US is switching slowly.

As a US citizen who is a metric fan and loves using it, in what way? The government did switch - in the 70s - according to its own statutes, it had to.

It has crept up in various places (and I find it hilarious) innoculously, like in 2 liters of soda, or in how computer processors are talked about in mm² die areas.

But the average American still uses imperial units religiously, anywhere they approach a problem involving any unit of measurement they always default to imperial, and having a 14 year old brother I see no change in his education or habits to indicate a slow transition of mindshare. The government moved decades ago, but the people aren't moving at all.

I get the impression it is much like high school language classes - you learn it once early on, never practice it, and by the time you are a full adult you have completely forgotten it. I'm not sure how to improve the situation to actually get the people to start using international standards, because if you were to start trying to force it on the supply side people would just not buy metric tools and information because they forgot it back in primary school.

I think the way to switch would be done the same way other countries have done so:

1. Make sure everyone is educated in metric

2. Change the easy things: the paper size the government uses, the units on food labels, the measures legal to use for sales of loose food or other goods, the units the government uses for all types of reporting. (Therefore if businesses want government contracts, they'll need to use metric.)

3. Change other standards, like residential construction, preferred fasteners, wire sizes. Where old measures are required for compatibility, write "24.5mm" in the standard. If the dimension could be changed to 25mm without any side effect, use that.

4. Change other things people see daily: I don't know if doctors use metric in the US, but I assume they communicate to patients in old units. Change the default, but accomodate older people. Change the road signs. Is anything left?

The UK is part way through 4, but has been stuck there for decades.

The current system is 800 years old, and over that time it has won over hundreds of different systems. A new system is proposed every now and then, and even though they might be better in a specific problem domain (say, microtonal music), but they always fall apart.

I have thought a lot about the problem (worked as a professional bassoon player for a very long time), and I can't say I have had many good ideas. There are some ideas for simplified music notation (with different shapes for flats and sharps) which work _very_ well for making sight reading easier. Until it doesn't: It can't express enharmonics (different ways of writing the same note), which makes tonality analysis harder, and can actually hamper readability since most people that are fluent in reading music usually "stay in key" when reading music.

A quick google gave me this: http://musicnotation.org/ and I can't say I am very impressed by anything I see there. But as you notice, most systems are oriented by lines. I don't think that is because people lack fantasy, but because it is a pretty good way to write music.

What do you think about parallel visualization? Right now, musical notation strives for a single notation that tries to encompass the entire work—and to also serve as a canonical, lossless transcription of the work, from which it can be recovered.

If you drop that requirement (and then assume digital storage) you could have 1. an underlying canonical format that has "all the information" but which is never presented to the performer, nor to the composer; and 2. a number of views that expose various dimensions of the composition. Like orthographic projections of a model in CAD software.

Presuming an interactive display (touchscreen, etc.) you could switch between these views at will; but even for printed sheet music, you could just isolate one measure at a time and then display several "stacked" views of that measure per page.

(Basically, picture widely-spaced, annotated sheet music, but where the annotations are themselves in the form of more musical notation, rather than words, appearing in additional sub-staffs attached to the measure.)

"Right now, musical notation strives for a single notation that tries to encompass the entire work—and to also serve as a canonical, lossless transcription of the work, from which it can be recovered."

I don't believe this to be true. (Modern) Guitar Music is most often written in tab often without accompanying staff notation. Also staff notation is not loseless, musicians will interpret the music differently. For example, with violin, whilst some instruction is given on bowing it is almost never complete and the musicians will find different ways to fit the bowing to the rhythm, this can make a huge different to overall tone as (most simply) the up bow sounds distinctly different to the down bow.

I do think this is the direction it is heading. There are new "smart" music stands coming to the market now with similar features.

Conductors can write notes about certain parts that can be accessed by musicians. Opera musicians (where different people play the same music every night) can have their own personal notes.

Most exciting is ofcourse that everyone has instant score access. That removes a shit-tonne of time wasted during rehearsals.

Those are just traditional use cases. I'm excited to see what will come. I don't know ifusic as it is practicedtoday can be "expanded" in any meaningful way, but that only time will tell

I think this is a great idea as part of a learning tool, being able to simultaneously visualise a musical idea on a score, in guitar TAB, woodwind fingering, piano roll etc.

I've got a plan on the backburner to do something like this using Ohm https://ohmlang.github.io/

Musicians frequently get taught music in large batches at schools, though, which means you don't have to worry about network effects—there are choke-points in the network.

There's no reason a given school couldn't teach a "colloquial notation" first, with the "Lingua Franca" musical notation taught later on, for everyone in that given school. Then everyone who comes from that school would know that colloquial notation.

Consider: the "Chicago school" of Economics; "Rugby School" football; etc. These things start as colloquialisms, then spread to global awareness.

England had a colloquial notation, taught in schools, for several decades: tonic sol-fa. But it could only talk about melody and rhythm, not harmony. It fell out of mainstream use in the late 1960s, perhaps as the music publishing industry consolidated and globalized, making it easier to have a single international edition of each song instead of separate editions by country.

For instance, while Imperial units aren't without some advantage, they are pretty generally inferior to the Metric system.

You say this pretty matter of factly, but I actually vehemently disagree. Many imperial measurements are better than their metric counterparts for day-to-day lay usage.

- Fahrenheit is a better scale than Celsius - Inches, Feet, & Miles are very practical units. Centimeters, and Meters much less so. - Pounds are smaller and offer better delineation than Kilograms. - Liters are pretty similar to quarts, though I admit the various Imperial sub-units are annoying.

Sure, it's easier to convert between metric scales, but the number of times I actually do that?: approximately zero.

Inches, Feet, & Miles are very practical units. Centimeters, and Meters much less so.

Really? Do you know how much easier it is to compute surfaces and volumes in metric systems compared to imperial? Concrete example. Figure how much soil you need to buy to fill a box knowing L, W and H. In metric it is a 10s process. In imperial i do not even know how you are supposed to do it. Does anybody even know how many quart are in a cubic foot?

> Does anybody even know how many quart are in a cubic foot?

They know it, after they do the conversion, via metric system.

(OK, nowadays you can just enter "1 quart to cubic feet" in Google. And the funnier ones you get at https://en.wikipedia.org/wiki/List_of_humorous_units_of_meas... )

No, I don't know the number of quarts in a cubic foot, but no one does because they're two different measurements for two completely different uses.

No wonder that you miss the point of metric units if you don't get why doing such transformations is useful.

In the metric system, converting between length, volume and weight is trivial and straightforward. This comes into play neatly whenever you need to pile up a precise amount of batter or liquid from containers measured with a different unit.

You don't really give any reason why Fahrenheit is a better scale or why inches, or why feet and miles are particularly practical units. It seems to me that people say this simply because they are used to them. You don't convert to metric units and it feels awkward because you don't use metric units.

There is an issue with "kilometer" being a complex word for everyday use (as compared to a mile) in the English language. That's more a linguistic issue that about the unit itself. Other languages have solutions to that with shorter colloquial name for the unit.

Of course the imperial units give a good opportunity for being funny, in ways like specifying speeds in furlongs per fortnight. But you can do the same in SI-derived units, like parsecs per picosecond.

> There is an issue with "kilometer" being a complex word for everyday use (as compared to a mile) in the English language. That's more a linguistic issue that about the unit itself. Other languages have solutions to that with shorter colloquial name for the unit.

Even in English people of a certain age can say "klicks" and be understood.

Exactly, it's something that mass usage will solve, even if the folk song will not sound just the same with "a hundred klicks, a hundred klicks, I am five hundred klicks away from home".

Other languages often say just letters "k" or "km".

“In metric, one milliliter of water occupies one cubic centimeter, weighs one gram, and requires one calorie1 of energy to heat up by one degree centigrade—which is 1 percent of the difference between its freezing point and its boiling point. An amount of hydrogen weighing the same amount has exactly one mole of atoms in it. Whereas in the American system, the answer to ‘How much energy does it take to boil a room-temperature gallon of water?’ is ‘Go fuck yourself,’ because you can’t directly relate any of those quantities.” Wild Thing by Josh Bazell.

That's great.

When is that EVER useful to the layperson?

Cooking

Another way to look at it is that the current notation system isn't the best overall, it's just the most tolerable trade-off between a bunch of mutually-incompatible requirements.

Replacing standard notation for all uses may be doomed to failure, but replacing standard notation for some particular use case (especially new use cases that weren't anticipated when standard notation settled into its current form) may be a very useful thing to do.

Computers also give us a few new options, such as displaying notation in a time-varying form, or using three dimensions, or notating the music in some universal language that isn't necessarily easy to read but that can be easily rendered in any desired notation.

Lattice notation for instance is something I really like, but I don't know how to represent it without some kind of animation.

Here's an example I stumbled across on Youtube awhile back of the kind of thing I mean: https://www.youtube.com/watch?v=jA1C9VFqJKo

Lattices generalize to higher dimensions, which means they might be amenable to virtual reality or even some sort of human-brain interface that allows you to experience 4 or 5 spacial dimensions at the same time.

> Another way to look at it is that the current notation system isn't the best overall, it's just the most tolerable trade-off between a bunch of mutually-incompatible requirements.

Isn't most tolerable trade-off between multually-incompatible requirements another way of saying "best overall"?

Totally agreed there are useful local overrides of standard notation. Tablature is one example, and there are others. I wouldn't call those replacements for standard notation though. Both notations exist, both serve different purposes, neither is going away, there's no either-or question to be resolved.

The lattice videos are super interesting! Thanks for sharing that. I want to watch a few more and understand his layout choices -- I think I kinda get it, triads form triangles. These don't encode anything temporal though, so this is a visualization that helps understand harmony spatially, but is not a musical notation and can't encode a song, right?

> Isn't most tolerable trade-off between multually-incompatible requirements another way of saying "best overall"?

I could have said that better. What I meant was that standard notation isn't better than every other system according to every metric we could use to compare such things.

Gary Garrett has more lattice demos on Youtube. Here's one that's an animation of an example in Harmonic Experience by W. A. Mathieu (which uses lattices extensively to explain harmony and is the best reference I know of for explaining how to understand them): https://www.youtube.com/watch?v=I49bj-X7fH0

A 3-5 lattice is a grid where one axis is fifths (powers of 3 in just intonation) and another axis is major thirds (powers of 5). Garrett implies a third axis for septimal flatted seventh (i.e. barbershop 7th) intervals. Since the grid is leaning to the right, the diagonals that lean the left are minor thirds. Powers of 2 (octaves) are usually ignored. Triangles that are flat on the bottom are major triads. Triangles that are flat on top are minor triads.

There isn't an obvious way to encode a whole song onto a single lattice diagram in a way that could be printed on a page and still be readable. They seem to work pretty well as animations or as static illustrations to explain chord transitions, though.

> What I meant was that standard notation isn't better than every other system according to every metric we could use to compare such things.

This is totally true; tablature is better for beginning guitar players to learn to play specific songs on the guitar.

The only reason tablature doesn't supplant standard notation is that the metric under which it's superior is much narrower -- it's only for guitars, and only better than standard notation for beginners.

I don't think standard notation is necessarily the best overall, by I do think it happens to be the best overall, the best we've got today. And I'm not convinced it will ever become a choice, as opposed to standard notation evolving like it has in the past to incorporate new ideas.

Thanks for the explanation of the lattic layouts; I hadn't noticed the triangle orientation part, I only got as far as seeing that horizontal lines formed the circle of fifths. I can't tell what the plus and minus symbols mean, do you know? Usually those are used for diminished and augmented chords and not single notes, so is Bb- another name for A that is useful under the lattice system?

It's a way to identify distinct pitches that are usually treated as the same in equal temperament.

For instance, in just intonation 2 (the major second of the scale) has a frequency that makes a ratio of 9/8 relative to the tonic, but sometimes you might want a slightly flatter major second with a ratio of 10/9. So, that note is label 2- to distinguish it from the regular major second.

Maybe no one has succeeded with a general replacement, but there are different notations for guitar. I assume some other instruments have their own notation too. When electric music kicked off, to reproduce sound you have to trade setups / circuit diagrams, old music notation can't encode that! I kind of think of it like x86 assembly. It's here to stay, for better or worse, but that doesn't mean you can't have nicer things on top, and there are still things that don't make any sense at all in the x86 world (like FPGAs for one).

Tablature has a long history as well, it didn't start with guitar. Before guitar there was lute and cittern tablature -- which typically use letters and not numbers. I play both guitar and lute and I actually wish the letters convention had stuck, it's more fun. Wikipedia says that the first known tablature was for an organ. https://en.wikipedia.org/wiki/Tablature#Origin

Yes, some other instruments have their own specific notations & tablatures as well. These aren't replacements for standard notation though, and never will be. They have a place, and they are useful, but they aren't in competition with standard notation. Tablature has its disadvantages (https://en.wikipedia.org/wiki/Tablature#Disadvantages) but also the single biggest reason for standard notation -- groups, band, ensemble & orchestral playing -- is something tablature can't help with at all.

Totally agreed that standard notation doesn't help with electronic sound reproduction, but I'd suggest that standard notation isn't for sound reproduction in the analog world either, that's not it's purpose. Standard notation is the sequencer, not the synthesizer. You can use standard notation to encode songs in the electronic music world, but it's definitely not super convenient, hardly anyone does that. The analog version of trading setups and circuit diagrams is carving your violin using plans and specifications of a Stradivarius violin.

I have a theory for this. Please, do not down vote me, I am here with limited English but really good intentions.

QWERTY keyboard is something humanity found a better solution, people have developed better layouts like Dvorak, per example, and world keeps using QWERTY (not my case).

I have studied long time ago that TCP protocol is also not the best protocol, there are much betters and faster, but people keeps using the old TCP for Internet...

I believe when something is already consolidated, it's expensive to change, sometimes it's not worthwhile do update all the consolidated knowledge/investment, even when having better solutions.

World updates consolidated solutions just when the gain really worth it, it's not the case for music notation.

I also agree with you, the music notation could be easier, but I believe they don't upgrade because the masters musicians have mastered it, so they like the actual notation, and they are the fellows with enough knowledge to create a better version. I believe there is others types of notation, but it would need to be used by the masters musicians, and music schools, and universities to start a wave that could replace the actual notation (that already works pretty well).

The argument that Dvorak is superior and that inertia is keeping people from converting has been studied, and while I think there's some element of truth, it doesn't seem particularly compelling, since big disruptive changes occur all the time.

"the best-documented experiments, as well as recent ergonomic studies, suggest little or no advantage for the Dvorak keyboard."

https://en.wikipedia.org/wiki/Dvorak_Simplified_Keyboard#Con...

"The trap constituted by an obsolete standard may be quite fragile. Because real-world situations present opportunities for agents to profit from changing to a superior standard, we cannot simply rely on an abstract model to conclude that an inferior standard has persisted. Such a claim demands empirical examination."

http://www.utdallas.edu/~liebowit/keys1.html

Musical notation is a vastly more complex system than keyboard layout, and I don't believe we have a Dvorak of music notation to even compare with. There are no contenders for musical notation that a large group of people believe are superior. So there's no reason to believe that inertia is keeping people from using another notation.

To go one step further, music notation is constantly changing, it has been evolving, adopting and incorporating the best ideas for thousands of years. What reason is there to not start with the assumption that it already took the best changes so far? I have no doubt that if superior ideas for notation develop in the next hundred years, that at the end of it, we'll still call the result 'standard music notation'.

My intention was not to compare music notation and dvorak, but write about human behavior in similar situation against the "inertia" you cited.

Totally, I understand. And mine wasn't to counter Dvorak specifically, but mention that the inertia theory has been questioned, and also mention that sometimes things are believed to be better by some people but in reality aren't much better if at all for most people. Sometimes inertia is posed as a reason for not changing when in fact the reason is the accepted system is the superior system for the largest number of people.

The latter is my theory about music notation; that inertia is not even at issue yet because there are no serious alternatives.

And inertia might never be an issue, because music notation is a fluidly changing system. TCP and qwerty/Dvorak are static systems that don't ever change, so you can argue about which one's better. Music notation is changing and improving, so it's hard to suggest that people are resisting change, and hard to suggest that something better will supplant it, right?

I agree with your theory in general though, outside of the issue of music notation, and I think a lot of people do. It's just a matter of finding the right examples that clearly demonstrate it. And it would be really interesting to somehow quantify the amount that something needs to be better before people will adopt it. It's like static friction in physics -- it takes more force to get something started moving than it does to keep it moving.

Now I got your point! Thank you.

That's something I've done time and time again, and seen others do too. It's easy to look at something and think you understand it well enough to know how it can be improved. But when you find out the rationale and reasons it is the way it is, it's kind of humbling. Like how it surprised me to learn that there's a lot of valid, practical reasons to use the Imperial measurement system over Metric.

> Like how it surprised me to learn that there's a lot of valid, practical reasons to use the Imperial measurement system over Metric.

I've always wondered about that. Why?

I don't remember where I read it, but one of the big reasons was that Imperial units are much easier to divide in ways that make a lot of sense in practical usage, whereas Metric is designed to make conversions easier for doing science, which puts practical usage on a lower priority. But take this with a grain of salt.

* 64.7989 mg of salt

Machinists and engineers often prefer Imperial "mils" (thousandths of an inch), for instance. It's easy to convert from kilograms to pounds or kilometers to miles, but there's no convenient metric unit for expressing typical distances and tolerances used in mainstream machining. A millimeter is way too coarse, a micrometer is way too fine.

As a specific case, in electrical work, it's easy for me to specify "6 mil trace/space" attributes for a PC board design. Not so easy to say "0.1524 mm" or "152.4 microns." If I round my specification down to 0.1 mm, the resulting copper features will carry less current and cost more money. If I round it up to 0.2 mm, other physical and/or electrical requirements won't be met. So now I have to add at least one more sig fig, which is a pain in the neck for no obvious benefit.

Also, what we have got this way through quite a bit of evolution...

The first thing that looks a bit like modern notation is probably plainchant, originating in the catholic church circa 14th centure:

https://i.ytimg.com/vi/4qKRGsJMMG8/hqdefault.jpg

The basic system we use today originates from about the 1600's or so, but has still evolved a lot.

There were tons of historical warts along the way that have largely dropped off - for instance, figured bass notation (https://en.wikipedia.org/wiki/Figured_bass) or the French violin clef (https://upload.wikimedia.org/wikipedia/commons/thumb/c/c3/Fr...)

See also: https://news.ycombinator.com/item?id=12159224 . Which I shamelessly plug since I was a participant in that one. :-)

I got a reply there that the current system is only suitable for professional musicians, and that you'd need something like shape notes to reach mass musical literacy. Now I'm hopelessly biased as a music degree-holder, a semi-professional musician, and a Presbyterian to boot ;-), but this strikes me as setting the bar way too low. Given levels of overall literacy in the US (which were very different when shape notes were developed) I don't think it's that difficult to learn the notation itself – the difficulty I think is in mastering the music system.

Reply on news.ycombinator.com