Thursday, July 21, 2011

The Acoustic Analysis of Laughter

Robert R. Provine has conducted some of the most fascinating studies on laughter.  From studies with Karen Emmorey observing vocal laughter and associated social variables in conversations between deaf signers using American Sign Language to a laughter study whereby he collected high-fidelity recordings of laughter (sans background noise) and used a sound analyzer to generate a detailed graphic and numerical description of sound frequency and amplitude. 

In Provine's study with deaf signers, they laughed at the same places in the stream of signed speech, at similar material, and showed the same gender patterns of laughter as hearing individuals did during vocal conversations.  Deaf individuals are very visual. They cannot follow long-winded dialogues and dramatic exchanges of verbal tones. Since they are visual, they're more of an "action" type of group. Rowan Atkinson (Mr. Bean) is hugely popular among deaf people given no explanation of his act is necessary.

In my post titled: The Golden Ratio of Laughter,  I discuss my own laugh study (smaller sampling) whereby I measured laughs at home. In my study, I found that a typical laugh climaxed at 1 with the .618 release (softer component) at the end, which is consistent with Provine's findings. A laugh 32 bars long would reach its peak in the 20th bar. (a + b is to a as a is to b").
I grew up watching Woody Woodpecker after school. Check out his laugh in this clip. "Ha, ha, ha, ha...ha" is consistent with my Golden Ratio of Laughter theory whereby the .3, .3, .3, .6 sequence is easily audible.  Each "bead" of the unit laugh (ha, ha, ha, ha...ha) demonstrates the three up and one slightly down laugh-unit consistent with The Golden Ratio of Laughter.

In Provine's study using a sound spectrograph, he and his colleagues quickly discovered the distinct acoustic signature of laughter, visualizing laughter as a series of evenly spaced sonic beads on a string. Each "bead" corresponded to a short, vowel-like laugh-note (syllable - i.e. "ha," "ho," "he") that had a duration (diameter) of about 1/15 second. The beads were spaced at regular intervals (onset to onset) of about 1/5 second. Typically, laughs proceeded with a descrendo, a gradual reduction in loudness as the laugh progressed. Or what I'd call: "Ha, ha, ha....ha." Not to be confused with "hee, haw..." 

From what we learn in Provine's study, it's difficult to stimulate a vigorous laugh when the laugh notes are are not homogeneous (he-he-he-he or ha-ha-ha-ha is easier than "he-ha-he-ha" or "ha-ho-ha-ho"). Variations in laugh-notes usually involve the first or last note in a sequence. thus, "cha-ha-ha" or "ha-ha-ho" are possible alternatives and almost as easy to produce as the homogeneous ones. 

Women's laughter usually has a higher fundamental frequency (female energie) than men's laughter (consistent with women having higher-pitched voices). The harmonics of women (and children's) laughter are more widely spaced than those of men because each is a multiple of a higher fundamental frequency. 

In an experiment conducted by Mike Cerri, director of the University of Maryland Baltimore County's recording studio, using a digital sound-editing and analysis system, Cerri removed the laugh-notes from a recorded laugh and closed the gaps previously occupied by the notes. Listening to the edited laughs, they discovered a long, unvoiced, breathy aspiration - a sigh. Normally, this sigh goes unnoticed because it's masked by the loud blasts of the adjacent laugh-notes. While the function of this sigh is still unknown, it is believed to be a by-product of the laugh-note production. Conversely, toward the end of a laugh, Provine noted that the decrescendo of normal laughter is softer (last note in the sequence is softer), which probably has something to do with running out of air.

The repertoire of laughs are fascinating (to me). The concept of examining giggles couldn't be funnier to to a lay investigator like myself. The short exhalation of breath chopped into staccato segments lasting about 1/15th of a second each and spaced 1/5th of a second apart is music to my ears. I download laughter and listen to it while reading jokes. While laughter may not cure the ills of the world, it sure makes the medicine go down smoother! 

Post a Comment