If someone screams, most people will run to their rescue, fearing the worst. But they may be surprised to learn that in fact the screamer got a promotion or won the lottery.
It could be equally hard to tell if someone is moaning from pain or pleasure, giving a roar of frustration or achievement, or a gasping in fright or surprise.
That’s because these emotive sounds use the same type of rough voice features called “nonlinear vocal phenomena”, according to a paper published in the journal Royal Society Open Science.
“Nonlinear vocal phenomena are commonly reported in animal calls and, increasingly, in human vocalisations,” write Andrey Anikin from Sweden’s Lund University and co-authors.
“These perpetually harsh and chaotic voice features function to attract attention and convey urgency, but they may also signal aversive states.”
Telling them apart is all about context, the study shows. Although the sounds tend to be strongly perceived in a negative light – and as more intense – when there are no cues, it seems they’re more likely to be interpreted correctly given adequate contextual details.
The researchers recruited 900 people for a series of three playback experiments to explore their perceptions of the nonverbal human vocalisations with positive, negative or no cues.
The synthetic sounds consisted of 48 human gasps, moans, roars and screams with ambiguous emotional qualities, derived from a database of 260 non-staged sounds and tested in a small pilot study.
In each condition, volunteers were asked to rate how the person making the sound felt on a sliding scale, ranging from “terrible” to “great”, and how intense the emotion was from “not at all” to “a lot”.
In the positive and negative conditions, participants were given visual and verbal cues. For instance, a moan would be accompanied by the words, “They are having a massage / sex / tasty food” or “They are in pain” and related pictures.
The authors suggest it’s rather surprising that people tended to perceive the vocalisations in a negative light as the default given the sounds can also reflect positive emotions in real life.
They note that animal communication likewise has positive examples of these rough, non-verbal sounds, such as affiliative calls by red wolves, songs of humpback whales and “pant-hoots” of chimpanzees.
Yet animals apparently also use contextual cues; for instance, bonobos make contest hoots in play as well as in conflict and their interpretation relies on the associated gestures, “which are softer in play”.
Without such contexts, the meaning of calls like screams or grunts can be “rather vague”, write Anikin and co, adding: “It is the listeners’ capacity for flexible, context-sensitive interpretation of such ambiguous signals that enriches their communicative potential.”
Even more obvious sounds like laughing can have less happy motives such as malice, sarcasm or nervousness, they note. “Thus, contextual cues are likely to enrich the meaning of most nonverbal displays, even those that are not inherently ambiguous.”