Musical components important for the Mozart K448 effect in epilepsy | Hacker News Reader

Musical components important for the Mozart K448 effect in epilepsy

nature.com

101 pointslnyan19 comments

Quality performance of K448:

https://www.youtube.com/watch?v=VIItKRaP2vc

https://www.youtube.com/watch?v=HVPgJ2f5bd4

https://www.youtube.com/watch?v=9iePyP2HOr8

This is also great, but I personally like more subtle and emotional rendition of Alicia de Larrocha / André Previn. In classical music performances the ultimate mastery is in the subtle differences, but as rule of thumb more emotional and human performances combined with clarity and precision are always better.

I tried to find a meantone tempered version of K448 but was unable to.

For those unaware, our contemporary tuning system for the piano, Equal Temperament, is a somewhat recent invention. It came after Mozart. There are dramatic differences if Mozart is played on the historically accurate Meantone tuning. For a sample see [0], although note that the cheerful final passage is believed to have been composed post mortem by someone else.

[0]: https://youtu.be/lzsEdK48CDY?t=700

Absolutely fascinating! Really curious what the underlying mechanism is.

Also found this bit particularly interesting:

> Apart from one other composition—Mozart’s Piano Sonata in C Major (K545)—the therapeutic properties of K448 could not be replicated with other musical stimuli. Stimuli previously tested were other Mozart compositions, Beethoven’s Fur Elise, and a string version of K448.

I had epileptic seizures due to a trauma. I felt them coming and learned how to stop them from occurring. A few years I was taking a medication, and every few weeks I had to deal with a feeling of a coming seizure. So I ended with a hypothesis that brain needs new stimuli to stop seizure. It needs something new coming, so it would be forced to change current activity from those crap it is trying to do now to something else. I was young and stupid, and my first seizure was in some way hand-made: a feeling of coming seizure was mysterious for me then, and I decided to explore it "scientifically". I learned how to strengthen it, and I was doing it for some time. The trick was to concentrate on any stimulus, to pick something "constant" from a sensory input and to shrink my attention to just that element of sensory input. Though "constant" it not in a physical way constant, in a mind's way: it feels constant, but in reality it is just highly predictable.

So my hypothesis about K448 is this is surprises at work. It somehow follow music laws, but you never know what happen next. Voices sometimes go along, then they go different ways. Is it one voice? Or two? Or five? Would it be trill in the end of phrase, or some straw arpeggio? When music will become loud, when soft? This music holds attention of listener and so it guides mind processes externally.

Though, it is just a wild guess based on an anecdotal data about epilepsy and single listening of K448. I invented "surpriseability" of a piece of music just now, and I didn't try to listen some other piece of music to compare its "surpriseability" to those of K448.

Thanks for sharing! How did you direct yourself away from seizing? Was there one moment where they stopped completely, or was it a gradual improvement?

I heard an anecdote about how the actor Danny Glover willed himself out of having seizures. Looking it up, epilepsy.com says they stopped "for reasons unknown", which seems interesting. I can see a community of people who haven't managed to escape from a negative bucket not wanting to believe it's possible.

I've also heard of research into brain implants that disrupt seizures before they start, which sounds similar to what you're saying (but more invasive).

My uncle has suffered from (trauma-related) epilepsy for a long time. He's been to lots of specialists but nothing has worked. I wonder if there's anything he can do mentally to grab onto the edges of the experience and divert it. IMO, when it comes to medicine, people can be too willing to trust the experts, and give up their own agency/internal feedback in the process.

It was a gradual improvement. These feelings (deja vu) became less sharp, less frequent, but more sticky, they were harder to shake them off. Then it stopped. I didn't finished a medication course, the doctor said I needed to take drugs for five years, I stopped to take them after a four years, because of side effects. But in a ten years or so it stopped completely. It is detectable in EEG -- I scared the shit from a teacher on a practical class while studying psychology: we learned about EEG, and so I was wearing EEG cap, sitting in a dark room with flashing lights, while my classmates watched my EEG, and she darted into the room, asking me how I feel, "is everything is ok?" It was ok, but she was scared by asymmetry in alpha-rhythms of my brain hemispheres, I saw it with my own eyes on the screen. So something detectable remains in my brains, but I feel nothing. For 10+ years.

> I can see a community of people who haven't managed to escape from a negative bucket not wanting to believe it's possible.

> I wonder if there's anything he can do mentally to grab onto the edges of the experience and divert it.

Yeah, someone told me that I really couldn't control seizures, because it is impossible, and therefore I just thought that I controlled them. Like the process goes by itself, and my conscious interpretation of them is an illusion of control. Something like rationalization. It made me to think. So I really do not know how it was. But then I've read Arnhild Lauveng "A Road Back from Schizophrenia", where the author describes how she became schizophrenic and then healed (an astonishing memoir, a look from the inside of the mind, that turns schizophrenic from normal and then somehow manages to get back, I believe you'd enjoy it), and how other people do not believe her, because schizophrenia cannot be healed, and I do not mind now whether my interpretation of my epilepsy is rationalization or not. It just is.

But in any case, scientifically speaking, I'm not sure that conscious efforts can help against an epilepsy: all I have is one data point, while brains is a very complex system, 1000 data points wouldn't be enough. But speaking as an error-prone human being I'm sure as hell that it could help. It might be difficult or even impossible to help patient from outside, because a psychotherapist cannot know how it feels from the inside in this particular case. Spoken language is not powerful enough to bridge the gap between minds, Lauveng shows this perfectly. Every individual need to discover her mind by herself and to find some levers to manage it. Maybe in my case the decisive thing was my "scientific" study of my deja vu before seizures started: I had a plenty of time to play with them. To strengthen it to a point when it became scary, to shake it off because I was too scared by it. I remember deja vu from my childhood (I had a trauma at ~3yo), though I became interested in them only in my teens.

It seems there may be some others...

From the wikipedia page:

> According to Epilepsy Action, research has suggested that Mozart's K 448 can have the "Mozart effect", in that listening to the piano sonata improved spatial reasoning skills and reduced the number of seizures in people with epilepsy. Apart from Piano Concerto No. 23, only one other piece of music has been found to have a similar effect.[2]

https://en.wikipedia.org/wiki/Sonata_for_Two_Pianos_(Mozart)...

If we follow the link we get this:

> Apart from another Mozart Concerto, K488, only one other piece of music has been found to have a similar effect, a song by the Greek-American singer Yanni. Entitled 'Acroyali/Standing In Motion', it is featured on his album "Live At The Acropolis". This composition was chosen by researchers because it was similar to Mozart's K448 in tempo, structure, melodic and harmonic consonance and predictability.

So, we have K448, K488, K545, and a Yanni song. (Piano concerto number 23 is K488.)

Probably there are others, it's just hard to test every song. (It'd be interesting to see a list of all the songs that were tried that were found not to have this effect.) It's also likely that there's a kind of music that would work even better, but it hasn't been composed yet. The search space of all possible music is very large, especially if you don't constrain yourself to 12-tone equal temperament or conventional time signatures.

I also wonder if it matters whether the test subject is familiar with the genre of music or not, and whether it matters whether the test subject actually likes the song or not.

Now I'd love to see that tested with Frank Zappa, for example with One Size Fits All.

> Stimuli previously tested were other Mozart compositions16, Beethoven’s Fur Elise19, and a string version of K44810.

Why would you not test other things too? Are there not enough of these IED episodes, or you have to wait around for too long for them to happen? Beethoven had piano sonatas, a lot of them, and Rachmaninov had some compositions for two pianos. There were composers around Mozart's time that wrote music in a similar style. Can people with these IED episodes test this theory for themselves using some music on their phones?

While this is cool, two critical points are (a) why no spectral analysis of K448 to determine what parts of the music are responsible for the effect and (b) IEDs are suggestive of the brain region primarily responsible for seizure onset but the jury is still out if they can be used as sole criteria.

If the emotional connection mentioned in the paper is correct, then the actual structure of the piece would dictate its effect, not a simple harmonic content -- although the Fourier Transform is bijective, so if you took into account details in phase I guess you're right.

I would doubt that a spectral analysis would be revealing, or indeed, of any use whatsoever. After all, such an analysis would vary from performance to performance, or even between recordings of the same performance. It would seem more likely that the musical contents would somehow be the relevant factor, although this is speculative, of course.

It’s probably not in the spectrum. It’s probably in those flowing mildly syncopated arpeggios.

I guess there's plenty of ways to tease apart the mechanism. I'd start by changing speed, try various frequency shaping (eg, squashing the treble), a version done in MIDI with pure tones rather than wavetable), try midi with wavetable, but with successively bigger changes to the tone, try playing various subsections on repeat, etc.

Yeah I've sometimes helped debug issues in the Rust compiler. People would come with a piece of code that made it crash or something. One task is to minify the example as much as possible. You start by commenting out as large components as possible, but need to watch out that nothing depends on the commented out code in a way that is detectable at compile time. So you remove piece after piece and finally end with a minimal example that reproduces the issue. In complex projects with lots of dependencies you often start out with large dependency trees and somehow have to reduce it. Sometimes it's easier to reduce something, sometimes it's harder. I guess here it's analogous when one wants to find the "core" part of the music that helps with epilepsy. Maybe one can try to vary it in a way that the effect is even larger.

Rhythms of theBrain

No, not music related, but frecuency. And maybe, waves.

Reply on news.ycombinator.com