Hacker News new | past | comments | ask | show | jobs | submit login
Why we love repetition in music (aeon.co)
169 points by adamzerner on Dec 25, 2014 | hide | past | favorite | 53 comments

As a musician I was prepared to roll my eyes at this article but it's really good.

When I was a kid I would play with Music Construction Set on my Apple ][, "painting" lots of random notes into one or two bars and then putting repeat signs around then. What sounded totally random and annoying at first became kind of catchy once you listened to it eight times in a row.

The composer that I really admire for unabashedly using repetition before the Minimalists really took off with it is Olivier Messiaen (the Turangalîla Symphony has lots of great examples). Classical music is historically largely about continual transformation, and he was unafraid to just cut and paste big sections. It worked magnificently.

I had the same reaction when I saw the title, but then I saw the author. Her book on repetition won the outstanding publication award from the Society for Music Theory this year. I haven't read it yet, but I've heard her present and she was really fascinating (even though I'm normally not interested in cognition at all).

I didn't realize she had a book on the subject (although of course now I see that it's mentioned in her byline). Thanks for calling my attention to it.

See also: Acid House. An astonishing number of classic tracks in the genre are based on "random" sequences on the Roland TB-303 synth, repeated hundreds of times. (The synth's sequencer is tenuous to program, so just pushing the buttons randomly is a quick way to get a sequence of notes)

Oh, but that filter, glide, and accent circuit can elevate. They sometimes behave so strangely.

The cut-and-paste aspect of Messiaen's music is definitely one of it's many unique features. That said, it's always been my belief that Messiaen got the idea from Stravinsky, specifically the Rite of Spring, which takes the same approach.

I heard somewhere - reference long lost, sorry - that repetitive music makes us happy, but slightly unpredictable repetitive music keeps us really happy.

It's like the brain has a threshold; under the repetitive threshold is boring, but too far above it is also boring. We need this middle ground where breakdowns or middle-eights or whatever they're called have this magical property.

This supposedly explains the popularity of jazz. Which leaves me personally baffled, but there you have it. The sciences!

The first thing I thought about while reading the article was minimalist classical music, like Philip Glass.

I think Glass does the repetitive/unexpected thing exceptionally well, though it's often subtle.

If you're already familiar with Glass (or if you like Beck) you might want to check out Beck's song from Rework, an album of remixed Glass: https://soundcloud.com/dunvagenmusic/nyc-73-78

It's repetitive, but has enough of the unexpected mixed in to make it a stimulating listen.

The second thing I thought about was Brian Eno and his attempt to 'learn' a sound recording of a city street by listening to the recording many times. He began to anticipate certain sounds and hear them as belonging to a structure.

I'm not finding a Web reference. It was described in a book by David Toop

The band Underworld has some old techno inspired by Glass. Dark, minimal. Very good if you like that kindof stuff. 'Second Toughest in the Infants' album-era.

I'd never heard of Rework_ before but I love Glass. Thanks for the link! In fact, I think this is now going to be a spur-of-the-moment extra christmas gift to my father.

This is great, thanks for sharing! Can you recommend more songs like that?

Jazz is built on very deep yet very simple structures - specifically, 4/4 swing rhythms, and circle-of-fifths harmony. I'll try to explain something of how swing rhythm works here, so you can grasp the depth of "slightly unpredictable repetitive" it offers, and the appeal of jazz (and its descendants, rock/country/R&B).

Swing is actually the simplest expression of a three-against-two polyrhythm. I don't know if this will format well on HN, but here's a try...



This is 12 subdivisions, divided into twos and threes. Where the twos and threes line up are interesting focal points. You can see them on the downbeat (start) and backbeat (middle). Now, let's try to express 3 against 2 in a lighter way...

| x--x-xx--x-x :||

or even


This last one is really interesting. The capital X (backbeat) is played with an accent to show a triplet-cycle lineup, and the trailing x is an "anticipation beat" into the next downbeat. With just those two things, you can express polyrhythm.

But unlike this notation, a beat isn't a fixed point in time! Rather, it's a region of probability where a sound might occur. Sounds can start anywhere in their beat, and attack or decay slowly or quickly, and evolve over multiple beats. This matters for the timekeeping instruments. Bass instruments tend to attack slowly, and high-pitched percussion like cymbals and snare drums attack very quickly. This creates a tension between the bass sense of time and the treble sense of time, tension that can be exploited musically.

So the backbeat and trailing anticipation beats that help define the underlying polyrhythm? They can be played early, or late. This can make the music feel faster or slower than it actually is!

Compare reggae and ska. They use exactly the same rhythmic structures. But in reggae, the guitar "skank" and the snare drum arrive very late, dragging the beat. In ska, they arrive early, pushing the beat. So reggae sounds relaxed, and ska sounds excited. Same beat, different meaning, just by manipulating the polyrhythmic emphasis.

The emphasis can also be manipulated by making the polyrhythm more or less explicit. Blues often has a very hard triplet feel. A lot of rock is very square and barely swings at all.

The beauty is that this is incredibly simple and repetitive, but offers an infinite variety of emotional expressiveness. This is why the American beat has dominated popular music for century. (As an aside, this is why so much drum machine music sounds painfully stiff - it lacks the core tool of expression)

This isn't a direct response to your points, but thank you for this! I'm a self-taught bassist, and I come from a very musical family, but I have no theory training at all. I love the form you've given to the things that I've come to just kind of intuit over the years of playing.

One of my favorite things to do on the bass is to play around with that "region of probability" - to jump the beat, or to wait until the last possible moment to resolve something before moving on with the riff. It's fun. It makes the music sound unpredictable-yet-familiar, alive, and exciting. When my drummer and I start playing around with this, really fun things start happening. I've never really known how to express it before, but this gives me a wonderful way to conceptualize it. Thank you!

How do these express a polyrhythm of 3 against two, exactly?

    Now, let's try to express 3 against 2 in a lighter way...
    | x--x-xx--x-x :||
    or even
I'm seeing the threes. The first is an alternating pattern of triplets:

    x-- x-x x-- x-x
And the second is four parts of near-repeat triplets.

    x-- x-- X-- x-x
So where are the twos in this polyrhythm?

There was a Nautilus article about this posted on HN recently [1], it's an interesting read about our mental expectations of musical patterns.

[1] http://nautil.us/issue/2/uncertainty/composing-your-thoughts

One of the most exciting things about Imogen Heap's music is that her songs tend to be repetitive enough to allow our brains to grasp the pattern, and just as we had done so, they take a left toin at Albacoikey and do something completely different.

Imogen Heap is a fucking genius.

Hide and Seek II off Songs for Tibet is my favorite thing: https://m.youtube.com/watch?v=T60Ttsb-n3w

Are there any songs you're referring to in particular?

"The Moment I Said It" is a good example.

>We need this middle ground where breakdowns or middle-eights or whatever they're called have this magical property.

It's called off-beat, If you mean rhythmic elements like the second base in the arch-typical Drum And Base, the and in One-and-Two-and-...

the third and to be precise

> repetitive music makes us happy, but slightly unpredictable repetitive music keeps us really happy.

Chopin's music does this really well. He's always changing the expected chord right at the last second to throw you off and then eventually comes back to it. There's a good TED talk about it.

If you are interested in this kind of psychological phenomena in music, check out book "Sweet Anticipation": http://mitpress.mit.edu/books/sweet-anticipation

I would guess that a broader idea applies to all fields of aesthetics. A thing is beautiful when it has the right combination of familiarity and surprise. It fits well with te idea of the brain as a pattern-matching machine.

Great article. It's wonderful to see rhythm celebrated for how it makes us feel.

The Guardian had an article a few days ago on an emerging emphasis of bridge rather than chorus, and the role of repetition, in recent blockbuster hits by Ariana Grande, Iggy Azalia, Nicki Minaj and others.

The article has a great quote from the founder of Motown:

         "Don’t bore us, get to the chorus!” Berry Gordy once joked, understanding that this is what listeners truly gather around.

This is a weird article. What is referred to as the bridge is the pre-chorus, a common element in pop music. Bridges are also common, but generally only happen once in a song (I can't think of any exceptions) and are typically a 'journey' to somewhere else before coming back to the original theme. An example of a bridge is about 2.5 minutes into John Mayer's Find Another You: https://www.youtube.com/watch?v=_QcaHBTfGU8&t=2m38s.

Amusingly, a lot of the tunes that are familiar as jazz standards, e.g., "Take the A Train," have a short AABA form, that the musicians refer to as the "bridge." But I suspect that jazzers abuse music terminology.

The bridge is where you find out who really knows the tunes.

Would you consider the Beatles' "I'm Looking Through You" to be a counterexample? Its bridge is performed twice.

This question might be slightly off-topic: some time ago I read an article (might have been on Wikipedia) which explained why music more than other art forms sticks with certain atmospheres, memories or moods. I unfortunately didn't bookmark it, and I couldn't find it again. Does anybody have an idea where to find it?

Fantastic article.

For those interested, certain artists like Swans[1], Godspeed You! Black Emperor[2], and ISIS[3] really take it to the next level with subtle changes in each refrain. As an added bonus, they are all fantastic for programming.

[1] https://www.youtube.com/watch?v=9HKIjG9JH6c

[2] https://www.youtube.com/watch?v=NivY_iRdSBQ

[3] https://www.youtube.com/watch?v=g3S3zuHRqlk

> this illusion demonstrates what it means to hear something musically. The ‘musicalisation’ shifts your attention from the meaning of the words to the contour of the passage (the patterns of high and low pitches) and its rhythms (the patterns of short and long durations), and even invites you to hum or tap along with it. In fact, part of what it means to listen to something musically is to participate imaginatively.

That's exactly how mysticism sees music. As a level above speech.

When I watch or listen to anything for the second time I only get about 75% of the enjoyment out of it. But it might be something wrong with me ...

Novelty effect is strong, but there might be a u-shape there. Have you ever tried to get past the valley?

I really enjoyed Spotify and had many play-lists with my favorite music. But when the discovery of new music declined and I got bored with my play-lists I closed the account and have never opened it again.

The curve never go up, but with some songs it have flattened and those are my favorites. But it has more to to with the feelings associated to those pieces then the music itself.

The sequence "sometimes behave so strangely" exhibits syncopated notes on a diatonic scale, 2/4 type meter, and a progression that adheres to chords of a melody. It seems unlikely that something this musical would often appear in normal speech, never before noticed.

Yeah, the sample was quite clearly handpicked for increased impact (and faster habituation, probably). But the argument is (I believe) about the fact that you can easily handpick such samples from perfectly normal spoken phrases.

Give me any person speaking any sentence and I'll make you a song.

Boots and pants and boots and pants and boots and pants

It's called – stutter edit.

> The stutter edit[1] is a musical production technique, most often known for its use in electronic music, in which fragments of audio are repeated in rhythmic intervals.

Electronic musician Brian Transeau (better known as 'BT') developed the "technique", coining the phrase, and later released it as a standalone plug-in.

[1]: http://en.wikipedia.org/wiki/Stutter_edit

No, it's not. Repetition has been with music since the cavemen.

Stutter edit is just a particular technique in working with samples, and not even that important in itself.

No it's not. A stutter edit is about much shorter pieces, and notably not about repetition. I'm not sure where Wikipedia got that from. It's about adding little silent gaps to tracks.

Adding silent gaps is closer to fine stacatto or granular synthesis. Stutter edit is taking very short slices of samples and rhythmically rearranging them at a very fast rate.

I'll agree that what the article is talking about and the stutter edit are very different things though.

Did you even read the wikipedia page? Check out "Skylarking" and "Tomahawk" by BT for a good sample of what it sounds like.

that wikipedia page, not to put too fine a point on it, is complete bollocks and reads like a marketing script for the izotope plugin.

check out the date on this announcement:


this program was "old" and wasn't the first. ableton added beat repeat in 2005, and there was vst's i can't remember the name of that did it before that.

you've hit a nerve anyway, BT is easily my least favourite electronic music artist ever and so I am probably not capable of reasonable debate hereafter. i think he's possibly the most overrated producer in the world.

I can pull examples of sample mangling that fall under these definitions from, at least, somewhere around 1993-1995. Someone seriously is saying BT invented this shit?

yeah people are seriously saying it. read that wikipedia page, it's beyond ridiculous.

earliest i can think of is the aphex remix of flow coma and i'm not even trying.

> Electronic musician Brian Transeau[3] developed the "technique", coining the phrase, and later released it as a standalone plug-in. Until this point the majority of stutter edits were created through deliberate manual editing techniques rather than automated processes (such as the eponymous plug-in). The audio plugin is named "Stutter Edit" and was co-released by iZotope and Sonik Architects.[4]

it's not like izotope release a plugin in 2002 which does the same thing, oh wait it is:


"developed the technique" is such a weaselly phrase. did he develop the technique, or did he further develop a technique that already existed. to my understanding neither is true. he's a crap musician to boot.

Transeau found that software tools to accomplish this weren't readily available. So, he decided to develop his own, forming his own software company, Sonik Architects. And later Sonik Architects was acquired by iZotope.

It's like saying: "Columbus didn't discover America because it already existed" or "Newton didn't discovered 'Law of inertia' because it already existed".

> Transeau found that software tools to accomplish this weren't readily available.

he could have used google, but keep believing!

Hava a source for that?

I'm pretty sure stutter edit is just one particular techinque used in various kinds of music, no so much about the more general definition of what makes music (= repetition) as the original article states.

I cringed reading that. There were a lot of people using that technique long before BT. Max/Msp, csound, and many other programs allowed user to do this. Not to mention the old amiga scene trackers and old jungle artists with their "trackers".

https://www.youtube.com/watch?v=6iiK4MgIPtI https://www.youtube.com/watch?v=_fX4qoruQik https://www.youtube.com/watch?v=I9BfB9HMWlc

It's a new technique (2011) that a certain sub-genre and neither software or the other artist made it back in past.

It was invented by BT and patented: https://www.google.com/patents/US8145496

Applications are open for YC Winter 2022

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact