Earlier this week we published an article about the new iPhone voice assistant, Siri. It was noted that in the U.S. and at least four other countries, Siri’s voice is female (France and the UK are the only two countries thus far that are featuring Siri with a male voice). A CNN article this morning presents a theory that posits that most voice recognition applications are powered by female voices.
As a company who works extensively with voice recognition applications and software, Plum can corroborate this supposition to an extent. A majority of the voice applications operating on the Plum VXML platform are indeed female. This of course begs the question, why? And in addition, why are some countries receiving Siri-loaded iPhones that have a male instead of a female voice?
One answer to this question may in fact be biological. A large number of scientific studies have shown that a majority of people find female voices more calming and soothing to listen to than males. Per Stanford Professor Clifford Nass, “It’s much easier to find a female voice that everyone likes than a male voice that everyone likes.”
Nass goes on to say that the reason for the human preference of female over male voices in many instances is potentially due to the fact that the female voice is the first one human’s hear while in the womb. Research has shown that fetuses will indeed react to the sound of their mother’s voice, but not to other female’s voices or to their father’s voice.
According to CNN, another possible reason people often prefer female over male voices is because they possess a unique pitch and quality that makes them easier to detect. This example dates all the way back to World War II, when women’s voices were utilized in the cockpits of fighter planes because they were distinct among the many male voices, and possessed a sound quality that was much easier to distinguish.
However, voice applications are by no means restricted to using only female voices. In some instances (like in the UK and France) a male voice has been chosen to power Siri. In the 1990s, BMW was forced to issue a recall on their navigation system because they had received so many complaints from German men refusing to take directions from a woman.
So how do tech companies choose what sex their voice application will speak in? For starters, most programs come with a variety of speech engines, allowing users to operate their programs specifically based on their preferences.
Depending on what TTS engine they are using (Nuance, Natural Voices, Cepstral), there are a variety of voices, accents and even languages users can choose from. Since Siri technology is brand new to the market, users don’t yet have the ability to change the default, but soon there will probably be a variety of voices and even accents users can select from.
While theories abound regarding Apple’s geographical gender choices for Siri, it is guaranteed that Apple spent a lot of time and money on market research and development to tailor the voice precisely to the market they were targeting. So while people’s evolutionary preferences may influence what type of voice they like to hear, this knowledge was extracted and the voice was assigned through the more scientific process of market research and subject observation and interaction.
