Sometimes behaves so strangely

by Kieran Healy on October 9, 2006

Just “listen to at least the first few minutes of this radio show”: (“or via links here”:, which begins with the work of “Diana Deutsch”:, a psychologist who studies the psychology of music. The opening segment demonstrates a remarkable phenomenon, whereby a looped segment of ordinary speech appears — after a few repetitions — to become musical. Moreover, once you’ve perceived it as music, listening to the segment in context makes it sound like the speaker is in a Busby Berkeley musical and has just begun to segue into a solo number. The general musicality of speech is obvious, I suppose, especially when you listen to certain accents, or hear “uptalk”: But this is a very nice sort of case.

Via Clifford at “Cosmic Variance”:

{ 1 trackback }

Jim » Blog Archive » I just found the soundtrack for my movie
10.10.06 at 8:42 am



Dennis 10.09.06 at 7:17 pm

Hey, that’s fantastic.


Orin Kerr 10.09.06 at 7:34 pm

The pianist Jason Moran has worked with this, including on his most recent album. He’ll loop a spoken phrase (sometimes in English, once in Turkish) and play the music of it on piano. Very cool.


Jon 10.09.06 at 8:38 pm

Steve Reich has been doing this for years. I was at a concert last week that included his “Come Out” from some time in the early to mid 1960s. But also see (hear) his Different Trains.


Jon 10.09.06 at 8:44 pm

Oh, and here’s Wikipedia on “Come Out”. Reich is quoted as saying “by not altering its pitch or timbre, one keeps the original emotional power that speech has while intensifying its melody and meaning through repetition and rhythm.”


Scott Spiegelberg 10.09.06 at 9:04 pm

Another instance of this is Alvin Lucier’s “I am sitting in a room.” He recorded himself speaking a paragraph while sitting in a room (naturally). Then he records the recording, played back in the same room. After many times through, the words are no longer intelligible, but specific pitches come out (no pun on Reich intended).


John B. 10.09.06 at 9:15 pm

Still another instance is Scott Johnson’s composition, “John Somebody.” Johnson adds instrumentation that plays, then plays off, the “melody” of the words.


Jay Conner 10.09.06 at 11:37 pm

As I watch my grandchildren develop the body language and expressions, and the tones and inflections long before they have words appropriate to what they are dancing and singing, it seems inherently obvious that language began as pantomime/charade, then moved to dance, then to music, and only later crystallized into language.

The one who is learning vietnamese from her mother and grandmother is already capable of distinguishing nuances of sound her father and I are unable to grasp. She knows music we can barely hear.


godoggo 10.10.06 at 12:26 am

Moran’s mp3 page has a fairly long sample of the Turkish one. It really sounds like she’s singing.


godoggo 10.10.06 at 1:03 am

(the “Turking one” I referred to is “ringing my phone”)


gdr 10.10.06 at 8:07 am

Different Trains makes the equation of speech and music explicit by playing a speech segment and then immediately repeating the same pitch contour on one of the instruments in the orchestra (but with the pitches adjusted slightly to fit the chromatic scale).

Reich’s The Cave has more of the same technique.


Ken Houghton 10.10.06 at 11:12 am

No one has mentioned Gavin Bryars yet? I feel old.


Ginger Yellow 10.10.06 at 12:13 pm

Does this work with an unstressed language like French?


gmoke 10.10.06 at 2:47 pm

Janacek notated conversation, the sea, and other sounds in daily life as he walked around listening. His music and his writing, he was also a music critic and essayist, are well worth exploring.

There is a theory that the Japanese language may make the natural world more intelligible to native speakers as it seems to be the only language that includes all the vowel sounds, both short and long, as words.


Z 10.10.06 at 2:54 pm

The whole program is fascinating and raises many very interesting cognitive science questions. I was instantly hooked by the sometimes behave so strangely tune (not only the tune by the way, but also the rhythm) but when I tried it with my french coworkers, the consensus reaction was along the line of “so what, it is a sentence repeated a few times, anyway English is always going up and down”. Incidentally, Diana Deutsch later explains that people speaking tone languages like Chinese have perfect pitch much more often than people speaking relatively monotonous language, which gave me some measure of comfort as my maternal language is French and I am practically tone death. What would happen if a sentence of French was repeated over and over? Probably no discernible tune would emerge but some rhythm might. I guess Aragon and Prévert thought so at least…


BrendanH 10.10.06 at 2:54 pm

“Musicality of certain accents”? Subtle self-reference there, perhaps?


godoggo 10.11.06 at 12:10 am

Almost makes me wish I hadn’t waited til 26 to start learning Chinese. Almost.

Chinese music is, overwhelmingly, horrible, horrible stuff. And don’t get me started on their singing.


bb 10.12.06 at 12:37 pm

Kieran, that was *wonderful.* Thank you!

The on-air demonstration was very cool! Not many media broadcasts can really *do* things to their audience like that.

But then they were too slow getting us to tonal languages. I thought I would be bored for the rest of the show until Diana Deutsch located perfect pitch there (in tonal languages). Woah! This is what people say. The notes speak to them. They have identities. Sweet. I was just amazed. It made me want to inhale my computer. Did I say thank you? Thank you.

Comments on this entry are closed.