Language Evolution

The more I read about mirror neurons the more unsure I become about their relevance. For the uninitiated, mirror neurons are a new type of neuron that has been discovered in macaques (and are presumed to be present in humans) that fire when observing someone performing an action. There is a nice little PBS mini-documentary on them here. Jim Hurford wrote an excellent article explaining their limitations (2003) and putting a deft boot into some of the more grandiose claims of the mirror neuron scientists.

So first lets look at some of those claims. Perhaps my uneasiness with mirror neurons is in part based upon the fact that I can’t help but feel there is something inherently pop sciencey about their media coverage. I don’t doubt their existence or the significance of their discovery, but they have emerged as a panacea concept, a functional missing link in our understanding of brain functions. A few people have argued that they explain everything in social interactions and language use, including the dominance of visual information in our species and the impact or imitation on behaviour. Rizzato and Arbib are the people who make the greatest claims about the relevance of mirror neurons. They hold the belief that:

“Human language (as well as some dyadic forms of primate communication) evolved from a basic mechanism that was not originally related to communication: the capacity to recognize actions.”

The ability to read others is here a precursor to the evolutionary motivation to communicate, a position that crosses over slightly with the Machiavellian Intelligence Hypothesis and Corballis’s theories on the gestural origins of communication. For Rizzato and Arbib the emergence of mirror neurons allowed us to develop sophisticated interactive behaviour that in turn lead to protolinguistic systems as ways of exploring the difference between action and perception. Mirror neurons were an essential step on the adaptive pathway drawn to full language use.

A slightly more nuanced analysis is presented by V. S. Ramachandran who argues that mirror neurons were more likely co-opted exaptations whose presence is necessary but not sufficient (to borrow those old philosophical terms) to explain the evolution of language. Mirror neurons permitted the first great leap towards tool use and culture because they permitted imitation through observation, and somewhere in this melting pot language evolution took place. The exaptation of mirror neurons is far more circumstantial than in Rizzato and Arbib’s account, and they are a self supporting explanation for language evolution.

Jim’s response to these claims is quite succinct and relies on several simple points:

1. Mirror neurons do not produce arbitrary signs – all mirror neurons permit us to do is feel the same stimulation from watching something as performing it ourselves. The abstract Saussurian sign that is essential to the formation of grammatical symbolic language could actually restricted by mirror neuron systems, because it restricts interations from making the leap to abstract symbolism.
2. Mirror neuron structures are probably common. Remember all these hypotheses about human mirror neuron activity are made based on experiments conducted upon macaques, not upon humans or even Great Apes – This is perhaps why my pop science detector was flashing. There must also be some sort of relationship between mirror neurons and motor neurons in order for the transferal from perception to action. To this extent it seems unlikely that mirror neurons are functionally separate from motor systems and Hurford argues that we will probably find a lot of mirror neurons in natural systems as it is the whole behaviour that is adaptive, not perception or action alone. Hurford also makes the case that many examples of mirror neuron stimulus are not necessarily caused by the individual living out an action through perception. It’s something like the difference between empathy and sympathy – we don’t always have to replicate first-hand experience through perception in order to experience a neural stimulation. The PBS documentary above argued that mirror neurons might explain the passionate connection a fan experiences through watching sport. But it’s not that the fan experiences victory vicariously through a mirror neuron-powered identification with their sporting hero, but instead that they experience a perceptual distance, a second degree sense of victory that lacks the same value as for the man on the pitch.
3. Finally, for mirror neurons to permit the mapping of symbolic meaning then signifiers and signifieds must fundamentally be the same. However, as is clear to anyone using a language, the many different aspects of symbolic language use, the word, the symbol, the phoneme, do not possess identical features. Mirror neurons facilitate the repetition of sameness, whereas linguistic systems are fundamentally built upon the linking of difference to the same function.

So yes, a very difficult subject, that still needs a lot of expansion if it is going to prove more than a sideshow. Like Jim I’m happy to conclude that “There is a long way to go from mirror neurons to language”. I’m dubious of their ability to explain that niggling evolutionary step from sociality to language, not least because they are present in creatures like Macaques that are social but not language-using. (Using the limited powers of Web of Knowledge and Google Scholar, I didn’t manage to find any research about mirror neurons in non-primates.) It seems you still need the sociality→mind reading→language pathway for them to play this vital role in language evolution. If we look at langauge as more an abstract complex skill than a mechanism for social cohesion, then mirror neurons seem much better suited to explain human empathy than human language.

If anyone can enlighten my skimmed knowledge of the subject I’d love to hear from you.