News article
Elephants can detect human foes by their voice and language
By: Jacqui Bealing
Last updated: Monday, 10 March 2014

Elephants bunch in response to Maasai playback. Credit: Dr Graeme Shannon

Professor Karen McComb and Dr Graeme Shannon. Credit: Katito Sayielel
Elephants are able to identify humans that pose a threat to them by distinguishing between the language and voices of different ethnic groups, according to new University of Sussex research published today (10 March 2014).
The study, carried out in Amboseli National Park in Kenya, involved family groups of African elephants being played sound recordings of the voices of two different human ethnic groups known to them: the Maasai, who, periodically come into conflict with elephants over access to water and grazing for their cattle, and the Kamba, whose more agricultural lifestyle poses less of a threat to elephants.
The results showed that elephants were more likely to demonstrate defensive behaviour, such as bunching together and investigative smelling, in response to male Maasai voices than male Kamba voices. Furthermore, their behaviour was also less defensive in response to voices of Maasai women and boys than to Maasai men, indicating that they also specifically take account of the sex and age of the voice to pinpoint the most threatening situations.
The ability to discriminate real from apparent threat, particularly in the case of human predators that differ in relatively subtle cues, has important impacts on the fitness of individuals as it avoids repeated interruptions to feeding and prevents unnecessary physiological stress.
Mammal communication expert Professor Karen McComb, lead author of the study published in Proceedings of the National Academy of Sciences (March 10 2014),says: “Recognising predators and judging the level of threat they pose is a crucial skill for many wild animals.
“Human predators present a particularly interesting challenge, as different groups of humans can represent dramatically different levels of danger to animals living around them.”
Previous studies have shown that African elephant family groups exhibit greater fear to the scent of garments worn by Maasai men than Kamba men, and also show aggression when presented with the red clothes that the Maasai typically wear.
Co-author Dr Graeme Shannon points out that acoustic cues, from which a herd can determine the ethnicity, gender and age of a potential predator, have an additional advantage in serving as an effective early warning system – especially if the predator is out of sight.
He says: “The human language is rich in acoustic cues. The ability to distinguish between Maasai and Kamba men delivering the same phrase in their own language suggests that elephants can discriminate between different languages. This apparently quite sophisticated skill would have to be learned through development or through younger family members following the lead of the herd’s matriarch and other older females.”
Professor McComb adds that the specificity of the elephant responses was particularly interesting. As well as attending to language cues, they were actually able to use fine-scaled voice differences to pinpoint whether the speaker was a Maasai man rather than a woman or boy who would be unlikely to cause harm. The study highlights the potential benefits of having these sorts of advanced skills for distinguishing between different subcategories within a single predator species.
Notes for editors
‘Elephants can determine ethnicity, gender and age from acoustic cues in human voices’, by Karen Mc Comb, Graeme Shannon, Katito N Sayialel and Cynthia Moss, is published in Proceedings of the National Academy of Sciences.
Professor Karen McComb and Dr Graeme Shannon are members of the Mammal Vocal Communication and Cognition Research group in the School of Psychology at the University of Sussex.
PNAS is one of the world's most-cited multidisciplinary scientific journals. It covers the biological, physical, and social sciences and mathematics and publishes cutting-edge research reports, commentaries, reviews, perspectives, colloquium papers, and actions of the Academy. PNAS appears weekly in print, and daily online as PNAS Early Edition. Newly published papers are listed at www.pnas.org/content/early/recent.For more information about PNAS or the NAS, visit www.pnas.org or www.nasonline.org.
University of Sussex Press Office: Jacqui Bealing and Maggie Clune. Email: press@sussex.ac.uk Tel: 01273 678888
View press releases online at: http://www.sussex.ac.uk/newsandevents/