News and Views

On the uniqueness of humankind: is language working memory the final piece that made us human?

The enhanced working memory (EWM) hypothesis

Recently, Frederick Coolidge and Thomas Wynn have developed an appealing hypothesis according to which a single additive genetic mutation might have increased working memory capacity in our species. As a consequence of this enhanced working memory (EWM), modern mind would emerge (Coolidge and Wynn, 2001, Coolidge and Wynn, 2005, Wynn and Coolidge, 2004). This change would have been modest and added to the abilities already possessed by pre-modern populations, who would nevertheless have

The central executive: nothing new under the sun

My comments about enhancement of the central executive capacity as the place where EWM could have occurred are in fact implicit in Coolidge and Wynn's proposal. These authors have already noticed that considering this component as part of working “memory” is in fact incidental, and that its description is far from different than the one used for executive functions. This has the advantage of being in line with many classical proposals on what “made us human” (Eccles, 1989, Russell, 1996,

The language processor

In contrast, the alternative of EWM as specific to the phonological loop is the most appealing and original suggestion in Coolidge and Wynn's proposal. But it is here where additional problems may lie. These authors confer to this slave system a steep leading role in the appearance of the syntax processes that convert our language in an outstanding attribute.

Indeed, Baddeley and Logie (1999; see also Baddeley, 2003) suggested that the phonological loop may be the bottleneck of language

The limited resources of the language processor and the number of active neurons

In several studies of event-related brain electrical potentials (ERP), it has been shown that good language comprehenders (either as measured by the Reading Span Test or by performance measures) display larger amplitudes of certain ERP components reflecting syntax processing than do bad comprehenders (e.g., King and Kutas, 1995, Vos et al., 2001). It has also been shown that these ERP components notably reduce in amplitude when parallel demands on certain linguistic working memory operations

Final comment: EWM specific only to the language processor?

If working memory were defined as the number of neurons that can be activated simultaneously, the dichotomy between the phonological loop (or, rather, the language processor) and the central executive as the locus for EWM as proposed by Coolidge and Wynn could be academic. Instead, EWM could have occurred in overall terms; that is, it was the number of neurons that can be activated simultaneously which was enhanced, regardless of whether these neurons belong to one or another partition within

Acknowledgements

The author wishes to thank Frederick Coolidge, Thomas Wynn, Juan Luis Arsuaga, and Diego Martín-Loeches for their helpful and valuable comments to improve the manuscript.