Sometimes behaves so strangely
September 27, 2019 5:13 PM Subscribe
Deutsch's 'Sometimes Behave So Strangely'
The Speech to Song Illusion.
The Speech-To-Song Illusion via 12tone.
The Speech-To-Song Illusion via 12tone.
Where do we draw the line between speech and song? It seems like a simple question: Singing sounds like singing, and speech... doesn't. But it turns out it's not quite that easy to nail down, because recent studies in music psychology indicate that the difference may not actually be sound-based at all. With a stunningly simple twist, researchers were reliably able to take clearly spoken audio samples and convince their listeners they were hearing singing. How?Video (and more interesting audio) found on the last link from 12tone's video notes: Diana Deutsch - Speech to Song Illusion
There's speech, there's song and then there's B-52 Fred Schneider. Hurry up and bring your juke box money.
posted by zaixfeep at 6:53 PM on September 27, 2019 [3 favorites]
posted by zaixfeep at 6:53 PM on September 27, 2019 [3 favorites]
Diana Deutsch's stuff always brought Art Of Noise's "Opus 4" into my head.
Also it was cool watching the look on the kids faces as they each "got it".
posted by not_on_display at 7:17 PM on September 27, 2019 [3 favorites]
Also it was cool watching the look on the kids faces as they each "got it".
posted by not_on_display at 7:17 PM on September 27, 2019 [3 favorites]
Fred Schneider's great, but the Sugarcubes' Þór Eldon is the king of the form:
I once met him,
It really surprised me,
He put me in a bath tub,
Made me squeaky clean,
Really clean.
posted by RobotVoodooPower at 7:53 PM on September 27, 2019 [2 favorites]
I once met him,
It really surprised me,
He put me in a bath tub,
Made me squeaky clean,
Really clean.
posted by RobotVoodooPower at 7:53 PM on September 27, 2019 [2 favorites]
Also, that Art Of Noise live performance clip is great because the original track is a studio construction, and they figured out how to do it live. Here's the studio track.
Also perhaps of interest: source poem, November.
Which honestly reminds me of Poetry For Two Voices, which when I encountered it in my 20s sparked a frenzy of scripting existing poetry for multiple voices, which led me to post these on the internet in a different (and yet somehow the same) lifetime.
posted by hippybear at 9:02 PM on September 27, 2019 [4 favorites]
Also perhaps of interest: source poem, November.
Which honestly reminds me of Poetry For Two Voices, which when I encountered it in my 20s sparked a frenzy of scripting existing poetry for multiple voices, which led me to post these on the internet in a different (and yet somehow the same) lifetime.
posted by hippybear at 9:02 PM on September 27, 2019 [4 favorites]
"What is song?" seems like one of those questions like "what is art?"
Someone can generate something godawful and someone will say, "well, that's definitely not art," but in likelihood, it is art, it's just Bad Art.
So too, singing is just speaking that sounds good. It's intentional with emphasis on pitch and timbre in addition to just the words (or sounds) that are said, but plenty of people have "speaking voices" with "musical qualities," and then again, plenty of us have "singing voices" which aren't much differentiated from our speech.
posted by explosion at 9:57 PM on September 27, 2019 [2 favorites]
Someone can generate something godawful and someone will say, "well, that's definitely not art," but in likelihood, it is art, it's just Bad Art.
So too, singing is just speaking that sounds good. It's intentional with emphasis on pitch and timbre in addition to just the words (or sounds) that are said, but plenty of people have "speaking voices" with "musical qualities," and then again, plenty of us have "singing voices" which aren't much differentiated from our speech.
posted by explosion at 9:57 PM on September 27, 2019 [2 favorites]
I'd also throw Falco into the mix. Y'know, just for discussion's sake.
posted by hippybear at 10:25 PM on September 27, 2019 [2 favorites]
posted by hippybear at 10:25 PM on September 27, 2019 [2 favorites]
I did not experience this illusion at all. I thought it sounded like speech all the way through.
However, there's quite a bit of evidence my brain isn't wired right.
posted by jamjam at 10:43 PM on September 27, 2019 [1 favorite]
However, there's quite a bit of evidence my brain isn't wired right.
posted by jamjam at 10:43 PM on September 27, 2019 [1 favorite]
this illusion is in line with what philosophers and musicians have been arguing for centuries, that strong linkages must exist between speech and music
You say tomato, I sing tomato.
posted by flabdablet at 2:56 AM on September 28, 2019 [2 favorites]
You say tomato, I sing tomato.
posted by flabdablet at 2:56 AM on September 28, 2019 [2 favorites]
singing is just speaking that sounds good
It doesn't even need to sound good. It just needs to sound like the cadence and rhythm and phrasing of it matters at least as much as the verbal content.
And since the essence of rhythm is repetition, it should be no surprise that music can be dragged out of damn near anything by repeating it often enough. God knows that's the only thing that keeps me from collapsing in despair every time I sit down behind the drum kit.
posted by flabdablet at 3:03 AM on September 28, 2019 [1 favorite]
It doesn't even need to sound good. It just needs to sound like the cadence and rhythm and phrasing of it matters at least as much as the verbal content.
And since the essence of rhythm is repetition, it should be no surprise that music can be dragged out of damn near anything by repeating it often enough. God knows that's the only thing that keeps me from collapsing in despair every time I sit down behind the drum kit.
posted by flabdablet at 3:03 AM on September 28, 2019 [1 favorite]
Oh, and it's also what can make just sitting on the deck on a cool spring evening and listening to the frogs jamming in the back yard such a transcendentally glorious way to spend a few hours.
posted by flabdablet at 3:06 AM on September 28, 2019 [2 favorites]
posted by flabdablet at 3:06 AM on September 28, 2019 [2 favorites]
This phenomenon is very familiar to fans of electronic dance music, which commonly samples and loops human vocalization of various sorts. It's not limited to speech, either – almost any sound can become musical if it's repeated, and presented in a musical context. What's that old jazz saying? Something like: if you play a wrong note, just play it again.
I've heard it hypothesized that singing – and, by extension, music in general – simply imitates and exaggerates the natural melodic and rhythmic cadence of human speech, and that this is why we respond to it. This seems plausible to me.
One can imagine that exaggerating one's vocalizations in this way might intensify a listener's emotional response, much as exaggerating one's secondary sex characteristics might intensify an observer's sexual attraction. Music (according to this hypothesis) is like adding MSG to normal speech.
posted by escape from the potato planet at 4:59 AM on September 28, 2019 [1 favorite]
I've heard it hypothesized that singing – and, by extension, music in general – simply imitates and exaggerates the natural melodic and rhythmic cadence of human speech, and that this is why we respond to it. This seems plausible to me.
One can imagine that exaggerating one's vocalizations in this way might intensify a listener's emotional response, much as exaggerating one's secondary sex characteristics might intensify an observer's sexual attraction. Music (according to this hypothesis) is like adding MSG to normal speech.
posted by escape from the potato planet at 4:59 AM on September 28, 2019 [1 favorite]
I'm old enough to remember being impressed by the release of Zoolook on the leading edge of that. Still got my vinyl copy of that somewhere.
posted by flabdablet at 5:50 AM on September 28, 2019
posted by flabdablet at 5:50 AM on September 28, 2019
reciting (not in song) the first four lines of “Mary Had a Little Lamb,” asking if it sounds normal, then reciting just the third line 10 times in a row.
I'm having trouble figuring out how you're dividing up the lines of "Mary Had a Little Lamb" so that there is a third line that's not identical to either the first or the second one.
posted by straight at 8:56 AM on September 28, 2019
I'm having trouble figuring out how you're dividing up the lines of "Mary Had a Little Lamb" so that there is a third line that's not identical to either the first or the second one.
posted by straight at 8:56 AM on September 28, 2019
I assume they mean this formulation:
Mary had a little lamb,
Her fleece was white as snow,
And everywhere that Mary went
Her lamb was sure to go.
Which highlights a point relevant to this discussion: The form of Mary Had a Little Lamb that is typically sung already incorporates rhythmic repetition.
posted by Mr.Encyclopedia at 9:24 AM on September 28, 2019 [1 favorite]
Mary had a little lamb,
Her fleece was white as snow,
And everywhere that Mary went
Her lamb was sure to go.
Which highlights a point relevant to this discussion: The form of Mary Had a Little Lamb that is typically sung already incorporates rhythmic repetition.
posted by Mr.Encyclopedia at 9:24 AM on September 28, 2019 [1 favorite]
It's well known that people who are aphasic after a stroke are often much less impaired in singing words, to the point that singing becomes the path to effective rehabilitation, and there are studies which seem to show that rhythm and cadence are more important to this ability than melody:
In short, repetition of a phrase causes the usual language centers in the left hemisphere to become fatigued, your brain then switches to the right hemisphere to handle the phrase, and that makes it seem musical because the right hemisphere generally deals with songs.
And this also would explain why I am not subject to this illusion, because when I had a left side subdural hematoma which was pressing directly on the regions of my brain which when damaged by strokes ordinarily make people aphasic, the effect on my language was not aphasia but much greater than normal verbal fluency. All fillers, hesitancies, and pauses disappeared, and instead of producing pigs of language, it came out of my mouth in an unbroken bar, and according to my partner was vivid, riddled with metaphor, and so uncensored that it alarmed people within earshot on multiple occasions. I do remember that during that time, I had absolutely no idea what I was saying until I heard it myself.
posted by jamjam at 12:19 PM on September 28, 2019 [3 favorites]
To find out whether and how singing works, Stahl and colleagues conducted a study in which 17 stroke patients with resulting non-fluent aphasia had to articulate several thousand syllables, which were sung and recited with rhythmic or arrhythmic accompaniment. The texts selected were linguistically similar but varied greatly in their level of familiarity and how formulaic they were.I think this is evidence for a hypothesis mentioned in one of the FPP links (my apologies for not remembering which) to the effect that repetition makes words lose their meaning for many people and this plays a role in the speech to song illusion, which in this circumstance we might translate into neurological terms as that repetition defeats or fatigues the left hemisphere and brings the right hemisphere into play, since people who are able to sing words after a stroke do seem to use their right hemispheres to sing their words.
The results showed that singing was not the decisive factor for the patients. Singing the texts did not produce better results than speaking them rhythmically. "The key element in our patients was, in fact, not the melody but the rhythm" says Stahl. The positive effect was greatest in patients where deeper brain areas, known as the basal ganglia, were affected. These areas are known to be crucial for rhythmic processing.
However, the level of familiarity with the song lyrics and whether the texts contained formulaic phrases was found to be even more important. Producing formulaic phrases and well-known song lyrics may involve other brain mechanisms than spontaneous speech, the researcher assumes. Daily expressions like 'How are you?' are highly automatized at the motor level, and common song lyrics can be recalled from long-term memory. In other words, formulaic phrases and familiar song lyrics may be easier for a patient to articulate -- regardless of whether they are sung or rhythmically spoken.
In short, repetition of a phrase causes the usual language centers in the left hemisphere to become fatigued, your brain then switches to the right hemisphere to handle the phrase, and that makes it seem musical because the right hemisphere generally deals with songs.
And this also would explain why I am not subject to this illusion, because when I had a left side subdural hematoma which was pressing directly on the regions of my brain which when damaged by strokes ordinarily make people aphasic, the effect on my language was not aphasia but much greater than normal verbal fluency. All fillers, hesitancies, and pauses disappeared, and instead of producing pigs of language, it came out of my mouth in an unbroken bar, and according to my partner was vivid, riddled with metaphor, and so uncensored that it alarmed people within earshot on multiple occasions. I do remember that during that time, I had absolutely no idea what I was saying until I heard it myself.
posted by jamjam at 12:19 PM on September 28, 2019 [3 favorites]
Okay, I had a chance to try this out loud and I feel like I recognize exactly what is going on from rehearsing music.
When you are learning a difficult piece of music, it's common to take one tricky phrase and work on it all by itself. First you make sure you can play each note correctly, then you play them very slowly in sequence over and over until you can do it right, then you keep repeating it, gradually getting faster until you can play it correctly at the right speed.
Then when you put it back into the music, when you get to your tricky passage, instead of thinking your way through each individual note, you say, "Oh, hello tricky passage!" and play it as whole phrase. It's a little bit like the difference between sounding out a word one letter at a time and recognizing an entire word. When I am practicing I can often consciously perceive the transition from recognizing and playing individual notes to recognizing and playing musical phrases.
When I tried the Mary Had a Little Lamb poem, on about the 4th or 5th repetition, I had what felt like the same sensation, noticing it go from individual words to a single phrase that I was recognizing by the shape of the pitches and rhythms. It felt like the same little rush of pleasure that I usually label "I am successfully learning a bit of music."
Even if (or especially if) you don't read music, but you can sing a familiar melody (say "Happy Birthday"), you're almost certainly not thinking, "Okay, the first two notes of "Happy" are the same, then you go up a little bit, and come back down to the same note again for "Birth-day", then jump up..." Your brain stores that in larger chunks. (Even in my example I couldn't resist grouping the pitches into words.)
So maybe the sensation of "bursting into song" is the feeling of recognizing the third line in the same way our brain usually recognizes music. I think this theory is supported by the phenomenon of being unable to remember lyrics unless you sing them. Because they're not stored as words, they're stored with the music as chunks of sound.
I think this theory is actually pretty congruent with jamjam's.
posted by straight at 4:36 PM on September 28, 2019 [2 favorites]
When you are learning a difficult piece of music, it's common to take one tricky phrase and work on it all by itself. First you make sure you can play each note correctly, then you play them very slowly in sequence over and over until you can do it right, then you keep repeating it, gradually getting faster until you can play it correctly at the right speed.
Then when you put it back into the music, when you get to your tricky passage, instead of thinking your way through each individual note, you say, "Oh, hello tricky passage!" and play it as whole phrase. It's a little bit like the difference between sounding out a word one letter at a time and recognizing an entire word. When I am practicing I can often consciously perceive the transition from recognizing and playing individual notes to recognizing and playing musical phrases.
When I tried the Mary Had a Little Lamb poem, on about the 4th or 5th repetition, I had what felt like the same sensation, noticing it go from individual words to a single phrase that I was recognizing by the shape of the pitches and rhythms. It felt like the same little rush of pleasure that I usually label "I am successfully learning a bit of music."
Even if (or especially if) you don't read music, but you can sing a familiar melody (say "Happy Birthday"), you're almost certainly not thinking, "Okay, the first two notes of "Happy" are the same, then you go up a little bit, and come back down to the same note again for "Birth-day", then jump up..." Your brain stores that in larger chunks. (Even in my example I couldn't resist grouping the pitches into words.)
So maybe the sensation of "bursting into song" is the feeling of recognizing the third line in the same way our brain usually recognizes music. I think this theory is supported by the phenomenon of being unable to remember lyrics unless you sing them. Because they're not stored as words, they're stored with the music as chunks of sound.
I think this theory is actually pretty congruent with jamjam's.
posted by straight at 4:36 PM on September 28, 2019 [2 favorites]
I'm looking for connections here to learning basic skills (I'm thinking about mathematic skills as I write this) slowly and painfully and then being able to see the music in them. I have a strong sense that I've read something insightful along those lines, but this morning's searches haven't proved fruitful for me.
I can't remember if it was presented as a left brain/right brain thing, or a system 1/system 2 thing or what. I see it a lot in students who are learning algebra or basic trigonometry. Basically following the steps blindly (and therefore blind to mistakes) morphing into 'playing' with shapes and equation forms. When I see that transition where they are the 'melody/rhythm', I know that they have cleared the biggest hurdle to understanding.
I'm really glad this was posted. I have more reading to do, I hope!
posted by Acari at 8:36 AM on September 29, 2019
I can't remember if it was presented as a left brain/right brain thing, or a system 1/system 2 thing or what. I see it a lot in students who are learning algebra or basic trigonometry. Basically following the steps blindly (and therefore blind to mistakes) morphing into 'playing' with shapes and equation forms. When I see that transition where they are the 'melody/rhythm', I know that they have cleared the biggest hurdle to understanding.
I'm really glad this was posted. I have more reading to do, I hope!
posted by Acari at 8:36 AM on September 29, 2019
I assume they mean this formulation:
Mary had a little lamb,
Her fleece was white as snow,
And everywhere that Mary went
Her lamb was sure to go.
Which highlights a point relevant to this discussion: The form of Mary Had a Little Lamb that is typically sung already incorporates rhythmic repetition.
Yes, exactly that. In my experience, the first, second, and forth lines continue to sound rhythmic, but after the repetition of the third line in isolation, that one alone sounds like singing.
So too, singing is just speaking that sounds good. It's intentional with emphasis on pitch and timbre in addition to just the words (or sounds) that are said, but plenty of people have "speaking voices" with "musical qualities," and then again, plenty of us have "singing voices" which aren't much differentiated from our speech.
One might say that singing is just sustained talking!
posted by LEGO Damashii at 5:07 PM on September 29, 2019
Mary had a little lamb,
Her fleece was white as snow,
And everywhere that Mary went
Her lamb was sure to go.
Which highlights a point relevant to this discussion: The form of Mary Had a Little Lamb that is typically sung already incorporates rhythmic repetition.
Yes, exactly that. In my experience, the first, second, and forth lines continue to sound rhythmic, but after the repetition of the third line in isolation, that one alone sounds like singing.
So too, singing is just speaking that sounds good. It's intentional with emphasis on pitch and timbre in addition to just the words (or sounds) that are said, but plenty of people have "speaking voices" with "musical qualities," and then again, plenty of us have "singing voices" which aren't much differentiated from our speech.
One might say that singing is just sustained talking!
posted by LEGO Damashii at 5:07 PM on September 29, 2019
I'm going to also submit as a blend of melody and speech Wham!'s song Everything She Wants. It's a song written nearly exactly as a voice saying the words would say them. It's brilliant.
posted by hippybear at 7:52 PM on September 29, 2019 [1 favorite]
posted by hippybear at 7:52 PM on September 29, 2019 [1 favorite]
It never works for me. In music with speech, and I listen to a lot of music like that, it still sounds like speech, and that's why I listen to it. I enjoy the little bits of spoken word in my music.
posted by evilDoug at 7:37 AM on October 1, 2019
posted by evilDoug at 7:37 AM on October 1, 2019
« Older Things change, literary edition | Red to Green Newer »
This thread has been archived and is closed to new comments
You can try it yourself with a friend (or group of friends,) by reciting (not in song) the first four lines of “Mary Had a Little Lamb,” asking if it sounds normal, then reciting just the third line 10 times in a row.
Then recite the entire poem again and watch everyone’s look of surprise when they perceive you bursting into song on the third line and then going back to speaking.
I love this illusion and Dianna Deutche is great. (Caveat: If you listen to that podcast, you will have that phrase stuck in your head as a song for the rest of your life.)
From the paper:
“To conclude, this illusion is in line with what philosophers and musicians have been arguing for centuries, that strong linkages must exist between speech and music.”
From The West Wing:
“Words. Words when spoken out loud for the sake of performance are music. They have rhythm and pitch and timbre and volume. These are the properties of music, and music has the ability to find us, and move us, and lift us up in ways that literal meaning can't."
—Aaron Sorkin, Writer; performed by Martin Sheen as President Jed Bartlet. S03E05
posted by LEGO Damashii at 6:42 PM on September 27, 2019 [8 favorites]