Why does symbolic communication in humans develop primarily in an oral medium, and how do theories of language origin explain this? Non-human primates, despite their ability to learn and use symbolic signs, do not develop symbols as in oral language. This partly owes to the lack of a direct cortico-motoneuron control of vocalizations in these species compared to humans. Yet such modality-related factors that can impinge on the rise of symbolic language are interpreted differently in two types of evolutionary storylines. (1) Some theories posit that symbolic language originated in a gestural modality, as in “sign languages.” However, this overlooks work on emerging sign and spoken languages showing that gestures and speech shape signs differently. (2) In modality-dependent theories, some emphasize the role of iconic sounds, though these lack the efficiency of arbitrary symbols. Other theorists suggest that ontogenesis serves to identify human-specific mechanisms underlying an evolutionary shift from pitch varying to orally modulated vocalizations (babble). This shift creates numerous oral features that can support efficient symbolic associations. We illustrate this principle using a sound-picture association task with 40 learners who hear words in an unfamiliar language (Mandarin) with and without a filtering of oral features. Symbolic associations arise more rapidly and accurately for sounds containing oral features compared to sounds bearing only pitch features, an effect also reported in experiments with infants. The results imply that, beyond a competence to learn and use symbols, the rise of symbolic language rests on the types of signs that a modality of expression affords.

Studies have revealed a link between a deficient ability to reproduce rhythm and dyslexia. Moreover, a normal ability to reproduce rhythm has been shown to correlate with reading speed. Some associate these findings to effects of neural oscillations and a visual parsing of text input (Vidyasagar, 2013). The present study aimed to clarify the properties of the visual stimuli that support a correlation between rhythm reproduction and reading speed. The experiments were partly based on Tierney and Kraus (2014). Thirty participants were asked to reproduce heard regular (2 beats/sec) and irregular (1-3 beats/sec) rhythms by tapping on a keyboard. Then, the participants had to read out loud, at a fast rate, visually displayed sequences of words and non-words. The sequences contained either no spaces between items (baseline condition) or spaces marking regular and irregular groups. Mean reading speeds were calculated on accurately decoded sequences. Among the significant findings, strong correlations were observed between mean reading speeds of texts containing spaces and the reproduction of certain rhythm patterns. However, no significant correlations appeared for text where there were no spaces separating words and non-words. This supports the view that the ability to reproduce rhythm can link to a visual parsing.