January 6, 2013
Speech Style Influences How We Perceive Gender Of The Speaker
redOrbit Staff & Wire Reports - Your Universe Online
According to a new study, the style of a person´s speech may help listeners guess their gender just as much as the high or low pitch of their voice. A researcher from the University of Colorado Boulder examined transgendered people in the middle of their sexual transformation to figure out how humans associate gender categories with different characteristics of speech.The study was based on the doctoral research of Lal Zimman, a PhD in linguistics at CU Boulder. While working on his dissertation, Zimman identified and tested two key features of speech besides pitch that humans use to perceive gender — the way people pronounce “s” sounds and the degree to which they intensify certain vocal sounds when they articulate (a feature known to linguists as ℠resonance´).
“In the past, gender differences in the voice have been understood, primarily, as a biological difference,” Zimman explained. “I really wanted to look at the potential for other factors, other than how testosterone lowers the voice, to affect how a person´s voice is perceived.”
Specifically, Zimman wanted to find out whether the style of a person´s speech affected how low their voice needed to drop before people began to perceive it as a male voice. In order to do this, Zimman recruited volunteers for his study who were undergoing a sex change from female to male.
One critical step in this process requires the patients to receive regular injections of testosterone, a steroid hormone that is produced naturally in both sexes but at levels 7-8 times higher in men. High testosterone levels are responsible for a number of characteristically male traits, one of which includes the deep tenor of the male voice.
What Zimman discovered was the voices of his participants could have a high pitch yet still be perceived as a male voice if the speaker pronounced their “s” sounds at a lower frequency than the average female. These lower-frequency “s” sounds can be produced by moving the tongue further away from the teeth when speaking.
What this indicates, say experts, is those characteristics we typically associate with male and female speech are not purely biological.
“A high-frequency ℠s´ has long been stereotypically associated with women´s speech, as well as gay men´s speech, yet there is no biological correlate to this association,” said Kira Hall, Zimman´s thesis adviser and Associate Professor of Linguistics and Anthropology at CU Boulder.
“The project illustrates the socio-biological complexity of pitch: the designation of a voice as more masculine or more feminine is importantly influenced by other ideologically charged speech traits that are socially, not biologically, driven.”
Zimman also found vocal resonance influenced the perceived gender of the speakers, although in a somewhat more nuanced manner than “s” sounds or pitch. Vocal resonance is the characteristic that causes a voice to seem like it is emanating from a person´s chest (as in most men) rather than their head (as in most women). Unlike “s” sounds, however, resonance is partly influenced by biology and partly by habit. For instance, vocal resonance is naturally lower when a person´s larynx is farther down in their throats. But when people are young, they also subconsciously learn to modify the position of their larynx when they speak, causing their vocal resonance to go higher or lower depending on whether they train their larynx to go up or down.
Testing the effects of “s” sounds and resonance on people´s gender perceptions proved to be a technical challenge. For the study, Zimman first made digital recordings of the voices of the 15 transgendered participants. In order to test the influence of “s” sounds, he used cutting-edge linguistic software to digitally modify the pitch of each recorded voice, sliding them up and down to more closely resemble the high and low pitches of female and male voices, respectively, while leaving the person´s characteristically female “s” sound in place.
Finally, he played the different voice recordings for a group of listeners and asked them to record whether they thought the voice belonged to a man or a woman. This allowed him to determine how low each participant´s voice had to go before most people started to recognize the voice as belonging to a male speaker.
What he observed was that those participants who made lower-frequency, masculine “s” sounds could still be recognized as male even if the pitch of their voice was in the high range typically associated with women.
Zimman´s results shed light on the complex interplay between biology and socially acquired traits — or, in pop-science parlance, nature vs. nurture — that is involved in how humans learn to associate gender with speech characteristics.
Zimman graduated in August and his research is currently being reviewed for publication. He presented his results on Saturday, January 5th at the 2013 annual meeting of the Linguistic Society of America in Boston.