Goat accents

In yesterday’s NYT Magazine, in the “one-page magazine” feature, this story about goat accents (“What a Well-Born Goat” by Hope Reeves):

It’s not just Eton alumni who distinguish themselves with their posh accents. According to a new University of London study, English pygmy goats (those farm-bred in Nottinghamshire, anyway) also display recognizable vocalization styles that morph as their social groups change. “It is not really a measure of animal intelligence,” says Alan McElligott, a co-author. “Nevertheless, the study does show a surprising additional cognitive capacity in a domestic animal that we are all very familiar with.”

In contrast to the great “cow dialect” story of August 2006, there’s real research here — McElligott leads a research group at Queen Mary University of London, “focussed on communication and cognition research, using goats, cattle and fallow deer” — but the little piece in the Times (with its fanciful dialect map of British goat bleats) frames the story in terms of large-scale dialect differences (by geographical region and social class) that will be familiar to its readers, though that’s not what McElligott’s research was about.

Here’s a New Scientist story of February 16th (“Young goats can develop distinct accents” by Andy Coghlan) that is on the whole more accurate, though the reference to accents is still potentially misleading:

Young goats learn new and distinctive bleating “accents” once they begin to socialise with other kids.

The discovery is a surprise because the sounds most mammals make were thought to be too primitive to allow subtle variations to emerge or be learned. The only known exceptions are humans, bats and cetaceans – although many birds, including songbirds, parrots and hummingbirds have legendary song-learning or mimicry abilities.

Now, goats have joined the club. “It’s the first ungulate to show evidence of this,” says Alan McElligott of Queen Mary, University of London.

McElligott and his colleague, Elodie Briefer, made the discovery using 23 newborn kids. To reduce the effect of genetics, all were born to the same father, but from several mothers, so the kids were a mixture of full siblings plus their half-brothers and sisters.

The researchers allowed the kids to stay close to their mothers, and recorded their bleats at the age of 1 week. Then, the 23 kids were split randomly into four separate “gangs” ranging from five to seven animals. When all the kids reached 5 weeks, their bleats were recorded again. “We had about 10 to 15 calls per kid to analyse,” says McElligott.

Some of the calls are clearly different to the human ear, but the full analysis picked out more subtle variations, based on 23 acoustic parameters. What emerged was that each kid gang had developed its own distinctive patois. “It probably helps with group cohesion,” says McElligott.

“People presumed this didn’t exist in most mammals, but hopefully now, they’ll check it out in others,” says McElligott. “It wouldn’t surprise me if it’s found in other ungulates and mammals.”

Erich Jarvis of Duke University Medical Center in Durham, North Carolina, says the results fit with an idea he has developed with colleague Gustavo Arriaga, arguing that vocal learning is a feature of many species.

“I would call this an example of limited vocal learning,” says Jarvis. “It involves small modifications to innately specified learning, as opposed to complex vocal learning which would involve imitation of entirely novel sounds.”

So we’re dealing here with relatively slight acoustic differences that have become shared in small social groups, presumably by a process of mutual accommodation (building on small individual differences in the bleats of the goats in a group) — referred to as call convergence in the Animal Behaviour article (abstract below). What’s crucial in this story is the mutual accommodation, which could, in principle, involve any bit of behavior that’s perceptible to the goats. There is an analogy here to linguistic features in small social groups (like cliques in high schools); such features tend to be shared within groups and to differ somewhat between groups. Neither linguists nor ordinary English speakers refer to these small group varieties as accents, much less dialects, but the (partially) systematic variation is nevertheless real, just on a small scale.

[Note on “cow dialects”, in two postings by Mark Liberman, “Oh, the moos you can moo” and “Where are moo from?”. The impetus for the news stories in that case came from a cheese manufacturer, who encouraged cheesemakers’ trade groups to publicize their regional varieties of cheese  — and swept distinguished phonetician John Wells into their campaign. Read the postings.]

[The abstract for “Social effects on vocal ontogeny in an ungulate, the goat, Capra hircus” by Elodie F. Briefer and Alan G. McElligott, Animal Behaviour 83.4.991-1000 (April 2012):

Vocal plasticity is the ability of an individual to modify its vocalizations according to its environment. Humans benefit from an extreme form of vocal plasticity, allowing us to produce a wide range of sounds. This capacity to modify sounds has been shown in three bird orders and in a few nonhuman mammal species, all characterized by complex vocal communication systems. In other mammals, there is no evidence for a social impact on vocal development. We investigated whether contact calls were affected by social environment and kinship during early ontogeny in goats, a highly vocal and social species. To test the influence of social environment on kid vocalizations, we compared half siblings raised in the same or different groups. The effect of kinship on calls was assessed by comparing full siblings with half siblings. Calls of half siblings were more similar when they had been raised in the same social group than in different groups, and converged with time. Full siblings had more similar calls than half siblings. The group-specific indicators in kid vocalizations show that goat call ontogeny is affected by their social environment. This suggests that vocal plasticity could be more widespread in mammals than previously believed, showing a possible early pathway in the evolution of vocal learning leading to human language.

The investigators looked at 23 vocal parameters in this study.]

Leave a Reply

%d bloggers like this: