Those of you who have read a few of the posts here at EvoAnth might notice a rather familiar pattern emerging: we wish to understand how an interesting aspect of our species evolved, but that aspect does not preserve well forcing people to develop rather ingenious work-arounds.
The reason this pattern has appeared across most of my posts is because it’s a pattern that appears across most of evolutionary anthropology. That’s because bones, whilst providing a lot of information on the past, do not answer every question we ask. Humanity’s ability for curiosity has always outpaced its ability to actually answer those questions.
Nowhere is this more apparent than when studying the evolution of speech. Being one of the cornerstones of our success, it’s a deeply interesting issue. With speech we can communicate complex ideas, allowing increasingly impressive feats of co-operation eventually culminating in the technology and civilisation that has enabled us to spread all over the planet.
At the same time it leaves behind very limited traces in the fossil record. The vocal tract, being fleshy, decays along with most of the other apparatus involved in talking. And lets not forget that speech itself doesn’t fossilise. Prior to the invention of writing, we have no clue what was said by anyone which makes tracking its evolution somewhat…problematic.
As such, most attempts to unlock the mysterious story of speech have to focus on the other fingerprints speech leaves behind, or even abandon looking at fossil evidence entirely.
For example, the Homo erectus vertebrae from Dmanisi seem to be identical to modern Homo sapiens and so it is argued they could’ve supported the repository muscles needed for speech which attach there. A Homo heidelbergensis/neanderthalensis cranium from Atapuerca has an ear canal adapted to hearing the frequency humans speak at, unlike chimps.
There’s also Robin Dunbar, of social brain hypothesis fame, who has used calculations of group size to try and work out when speech became a necesity to maintain group cohesion. It’s all frightfully ingenious stuff that allows us to get a clearer picture of the development of speech than we would have had if people had simply given up when direct evidence disappeared.
And now new research has come out to join this rag-tag bunch of heroic papers, chipping away at the past despite having all the odds stacked against them.
This work is based around the fact humans don’t have air sacs in their vocal tract, whilst every other ape species does. Parsimony would thus suggest that lacking airsacks is an evolved feature in humans, rather than every other ape happening to develop them after they had split off the lineage that led to humans.
The exact function of these air sacs is unknown, but the scientists in this study don’t really care about what they do. It’s what happens when they disappear that they’re interested in.
Now, here we run into the classic EvoAnth problem: how to study this? Individuals from the past have lost their air sacs, even if they had them to begin with, which makes understanding how this changes speech ability quite difficult.
And one can’t simply compare the speech abilities of modern humans and chimps since there are a lot of other differences between them, how do we know which are the product of air sac absence? What they needed was something where everything was the same, they could just attach or remove air sacs to see what happens; something they could control the variables for….
What they needed was a model.
They then blew a bunch of air through all these models to see what would happen. The results? Removing air sacs lowered the frequency at which loudest sounds were made, increased the power of those sounds and moved them apart on the frequency spectrum so each sounded more distinct.
They then hypothesised that this might have a benefit to speech by allowing the wide range of sounds needed to be articulated more clearly by making them sound more different. It would be difficult to hold a complex conversation in a language where everything sounded the same, after all.
So they made their models make the vowel noises and then got people to try and identify which vowels were being said, the idea being that if air sacs make things less clear then people would misidentify the vowels made by those models more often.
And low and behold, they did.
Now there are some problems with this part of the study, such as small sample size, the most eyebrow raising being that we’re asking people. People who’ve spent their whole life attempting to identify these sounds and so are exceptionally good at it, which might amplify an otherwise small effect size.
But they do concede these problems and admit this is merely a preliminary study and they hope to solve this issues in the future. I look forward to when they do, but in the mean time the more solid data regarding absolute frequencies is robust enough to suggest the loss of air sacs is rather importance in the evolution of speech.
Or since we’re talking about the loss of something, perhaps the operative term should be devolution?
Oh, and the most exciting thing about this work? Whether an organism had air sacs or not is detectable in the fossil record. That’s right, we can go start studying the evolution of language in the long dead.
EvoAnth: 1, decomposition: 0
|de Boer, B. (2012). Loss of air sacs improved hominin speech abilities Journal of Human Evolution, 62 (1), 1-6 DOI: 10.1016/j.jhevol.2011.07.007|