ASA Lay Language Papers

2nd Pan-American/Iberian Meeting on Acoustics


Laughing with your Mouth Open Sounds Best

 

 

Michael J. Owren - owren@gsu.edu

Department of Psychology

Georgia State University

Atlanta, GA 30303, USA.

 

Tobias Riede – t.riede@utah.edu

National Center for Voice and Speech

University of Utah

Salt Lake City, UT 84101, USA.

 

Popular version of paper 5aSC211

Presented Friday morning, Nov 19, 2010

2nd Pan-American/Iberian Meeting on Acoustics, Cancun, Mexico

 

 

Laughter is a central part of human social interaction, starting in early infancy and continuing throughout the life-course. Both producing and hearing laughter are most often associated with positive emotion, although not uniformly so. For instance, recent research has shown that vowel-like, “voiced” laughter is consistently rated as a positive sound by listeners tested in laboratory settings. Noisier, “unvoiced” laughter is another very common form, however, and has been found to have little emotional impact. This finding is particularly intriguing given that both types are routinely produced by laughers experiencing positive emotion. In other words, all laughter from positive, enjoyable circumstances does not sound uniformly  positive to listeners, although it is not yet clear why not. The current work goes a step further in asking whether having the mouth open or closed while laughing also plays a role in the perception of these sounds, now examining voiced laughter in particular. Results confirmed that all voiced laughs are inherently positive to listeners, but also show that laughter produced with the mouth open has a significantly more positive impact on listeners.

 

Historically, there has been surprisingly little study of the human behavior of laughing, with most work instead focusing on topics such as humor as a common trigger of laughter. More recently, however, laughter itself has begun to receive much-needed attention, for instance addressing basic questions such as what laughter acoustics are like and those acoustics affect listener perception. Though much remains to be learned, one well-established finding is that laughter by itself can affect a listener’s emotional state, as demonstrated by the successful use of laugh tracks to increase enjoyment of television comedy programs. Even cursory listening to such tracks nonetheless reveals that they overwhelmingly emphasize one particular kind of laughter, namely the “open-mouth, voiced” version. Voicing simply means that the vocal folds are vibrating, whereas laughter can also occur in an unvoiced, “breathy” fashion in which air moves forcefully through the vocal tract but without engaging the vocal folds. This contrast is illustrated in Fig. 1, along with sound examples. The difference is intriguing because although both kinds of sounds are associated with laughter produced while in a positive state, laboratory studies have shown that only voiced laughter reliably induces positive emotion in listeners. Unvoiced laughter typically leaves listeners unmoved.

 

owren01.jpg

 

Click here to listen to an example of a voiced female laugh.

 

Click here to listen to an example of an unvoiced female laugh.

 

The current work took these observations a step further, comparing responses to voiced laughter produced with the mouth open versus having the mouth closed (see Fig. 2 and associated sound examples). Given that both kinds of sounds are voiced, one could anticipate the two to elicit similar responses. For example, that outcome would be expected if voicing is a marker of positive emotion occurring in the vocalizer. On the other hand, having the mouth open or closed could reflect higher or lower arousal levels in the laugher, a cue that listeners could also be unconsciously responding to.

 

owren02.jpg

 

Click here to listen an example of male open-mouth laughter.

 

Click here to listen to an example of male closed-mouth laughter.

 

In the current experiments, listeners heard a series of laughter bouts presented in random order over headphones, with each bout including at least one, and usually multiple, individual voiced bursts. The task in each case was to rate how positive each stimulus was, using scales that ran either from “very negative” to “very positive” or from “neutral or negative” to “very positive.” These sounds had been produced by laughers watching video clips or interacting socially, and were all associated with positive emotion. In the first experiment, 28 male and female university students rated 48 different laughter bouts, half being open-mouth sounds and half being closed-mouth, with half of each kind coming from male laughers and half from females. All the laughter was rated positively, although as in previous studies, female sounds were rated more positively than male versions. That result is illustrated shown in Fig. 3, along with the new finding that for both sexes, having the mouth open was also important, with open-mouth laughs being rated significantly more positively than closed-mouth versions.

 

owren03.jpg

 

Two other procedures tested participants when hearing only female laughter. One repeated the rating of open- versus closed-mouth laughter with new bouts, again showing open-mouth laughs to be significantly more positive. The last comparison tested for an effect of duration, contrasting rating outcomes for both voiced and unvoiced laughter with bouts that were either shorter (1 to 2 bursts) or longer (3 or more bursts). The results are shown in Fig. 4, where ratings of unvoiced as well as closed-mouth laughter were essentially unaffected. However, open-mouth, voiced laughter stood out in showing a clear effect of sounding more positive at longer durations. This result can be interpreted as indicative of a corresponding increase in the positive emotion being experienced by the listener when rating a given sound.

 

aowren04.jpg

 

In the context of earlier work, the current outcomes show that listener responses to laughter can range from neutral (unvoiced) to somewhat positive (closed-mouth voiced) to clearly positive (open-mouth voiced) sounds. Laughter with vowel-like, voiced bursts produced with the mouth open clearly stand out from other variants, with longer bouts of such laughter eliciting higher ratings and likely stronger emotional reactions in listeners. Here it is important to again note that all the laughter tested in the experiments had originally been recorded from vocalizers who themselves were experiencing positive emotions. One likely interpretation is therefore that acoustic differences in positively toned laughter are correlated with changes in vocalizer arousal level. In other words, voiced, open-mouth laughter may in particular stand out because as laugher arousal goes up, the vocal folds are increasingly set in motion and the mouth opens during sound production. Listeners may or may not understand that connection at a conscious level, but may respond emotionally to these changes through unconscious processing. Work is now underway to test the connection between vocalizer state and acoustics more directly by recording laughter from vocalizers who are reporting enjoyment level during the laughter-inducing task itself, with experimenters simultaneously measuring key indicators of their reactions at a purely physiological level.