Imaging language evolution

A comparative neuroimaging study performed by researchers from Emory University in Atlanta, Georgia, in collaboration with colleagues from the University of Oxford, provides clues to how human language evolved.

In the past, it was believed that the increase in brain size during human evolution occured mainly to accomodate our complex linguistic abilities. But the findings of this new study suggest that the emergence of language also required major modifications in how the brain is wired.

Using diffusion tensor imaging (DTI, a type of functional magnetic resonance imaging which I described in this post about synaesthesia), Rilling et al compared the structure of arcuate fasciculus, a large white matter tract, in humans, chimpanzees and macaques.

It is believed that the organization and terminations of the arcuate fasciculus of humans and macaques differ somewhat. However, the fibre tract in each of the species has until now been examined using different techniques (DTI in humans and axonal tracing in macaques), and so a precise comparison has not been possible. In the current study, Rilling et al have, for the first time, compared the arcuate fasciculus of humans, chimpanzees and macaques, using the same method.

In humans, this bundle of nerve fibres connects two regions of the cerebral cortex that are vital for language. From diffuse regions of the frontal lobe, including Broca's area, which is involved in speech production, the arcuate fasciculus projects backwards, arching around the Sylvian fissure, the prominent cleft which separates the frontal and temporal lobes. It then descends deep into the temporal lobe, where it branches and terminates in several different areas, including Wernicke's area, which is known to process the meaning of words.

untitled.bmp

By contrast, the connections between frontal and temproal lobes in the chimp and macaque are much weaker, as illustrated in the schematic diagram above. In chimps, the arcuate fasciculus branches less extensively in the temporal lobe, and makes far fewer connections there. And in macaques, the fronto-temporal connections are weaker still; the fibre tract barely reaches the temporal lobe, and does not form any branches.

This study therefore provides some evidence that the organization of the arcuate fasciculus was strongly modified in the human lineage, and that this was one of the brain specializations that led to the emergence of language.

In the macaque, the brain regions analogous to Wernicke's area are involved in higher order processing of visual information. It seems, then, that when humans diverged from other primates during the course of evolution, another specialization necessary for language was the addition of new cortical fields such as Wernicke's area.

This would in turn have necesitated a reorganization and enlargement of the arcuate fasciculus (leading to a disproportionate increase in the white matter volume of the frontal and temporal lobes), and would have displaced the visual cortical areas towards the back of the brain.


Rilling, J. K., et al. (2008). The evolution of the arcuate fasciculus revealed with comparative DTI. Nat. Neurosci. doi: 10.1038/nn2072. [Abstract]

More like this

That's fascinating, thanks for the post. Very minor correction: DTI is a structural, rather than a functional, imaging modality.

This, as I understand it, isn't contradicting the view that language could have come about to support cooperation when group size increased, is it? Might it not be that the branches of arcuate fasciculus grew more extensive as a response to more elaborate communication endeavours and not the other way around?

I wonder if the Bonobos would be closer to us, or more like the other chimpazees. I couldn't access the original article, but I think those chimpanzees were Common Chimpanzees, right?

One would expect that both since both sets of apes lack human language, both would have impoverished AF, but this is not the case. It makes sense if one defines 'language' non-technically to include any symbolic communication system, since chimps communicative abilities are much greater than macaques.

Also, Wernicke's area is specialized to process specifically linguistic information, whether the input is visual as in ASL or aural as in French. When humans developed new cortical fields to process specifically linguistic information, they recruited an area previously used for visual information, rather than auditory, and this supports a gestural origin for language.

But which came first? Did talking 'grow' the brain. Did talking provide an advantage even over seeing? Another puzzle too for those exploring our garbage DNA, eh?

What about the view that language faculty is hung upon primitive planning operations such as motor control and mirror neurones in the T5 of primates (homologous to Broca's area in humans)? If we look at primates that have learnt ASL, we can see that it sports autistic-like flaws, like a lack of spontaneity. Also, studies with chimps by Köhler (1925) showed that while primates can use tools to attain goals, such as stacking up crates to reach some bananas, they need to have the tools within reach, i.e., in the same room as the bananas. This shows that their plan-making ability is forward-chaining and reactive, which is quite appropriate for animals. Us humans on the other hand can do backward chaining, that is, work back from a goal to the initial steps of a plan.

Language cannot have sprung from improved acoustic processing, nor it could have improved from gestures. Gestures must have been the first vessel of language. But its real origin needs an increased ability for plan formation. Syntax and semantics are transparent to planning

(please excuse the excessively computational lingo i have used, i am after all, a student of cognitive science)

Please don't say apes have learnt ASL. Actual users of ASL are understandably sensitive about implications their language is so primitive even an ape can learn it. Apes have learned sign systems based on, or derived from, ASL, to be accurate.

What does it mean 'syntax and semantics are transparent to planning'?. Does this sort of thing show up in MRI studies?

Great article.

"But which came first? Did talking 'grow' the brain. "?

Why not?

Suppose that in infancy the teaching of speech by the mother wires the brain.

That would mean that we start off with a chimp-type brain. Under the influence of learning to speak that plastic mass assumes a configuration which allows us to "think with words". That is how we do it: we pluck words from our brains and string them into thoughts.

Birds teach chicks to sing in the same way that humans teach babies to talk.

The mysterious "key" to human evolution is when our particular mammalian species started using vocalization like the birds. That point was when we began to use vocalization for male courtship of the female. In other placental mammals, including chimps, females don't get courted because they go into heat and accept the nearest male for mating.

Picture our "Eve"; an H.erectus female born without estrus. When she ovulated she wasn't thrust into having sex. Rather, some male sitting near her was waiting for her to become interested. He made a noise. She looked up. He repeated the noise. Eve smiled.

At that point the process of humanization began.

arne@prontomail.com