Dan Dediu

Universitat de Barcelona

For a phonetician, all speech sounds are equally beautiful and fascinating. But some are more equal than others: the alveolar trill, or the trilled, vibrant or “Spanishr, is one of these. First, it is a complex sound from articulatory, motor and aerodynamic points of view, as it needs precise tongue positioning and stiffness and air pressure to get the tip of the tongue vibrating, rapidly closing and opening the airflow a precise number of times, to produce that lovely rrrroling sound of Spanish perrro or of Braveheart’s tirrany. Second, it is very hard to learn, not just by struggling second language learners, but also by children acquiring their native language, such as Spanish, Catalan or Romanian: it is one of the last sounds to be mastered by children, a few need speech therapy, and some, such as myself, never master it. So, why do so many languages have it? Shouldn’t it be rare, with most languages using “easier” kinds of r, of which there is no shortage?

Or is it? In work with Rémi Anselme (a bilingual French-Spanish who can’t roll his r’s) and François Pellegrino (who can, for fun), painstakingly going back to the data in the flagship journal for describing speech sound systems, the Journal of the International Phonetic Association (association that curates the International Phonetic Alphabet), we show that this is mostly an illusion. Many times, the written symbol “r” may not necessarily represent the trill r (whose phonetic symbol is also “r”) but may instead be used because it is easy to write (easier than say, ɾ, ɹ, ɻ or ʁ), because linguists expect it in a given language (neighboring or related languages supposedly have it), or because of confusing transcription guidelines.

In short, there’s much less trilling going on, defusing a linguistic paradox. Moreover, nothing is easy in language, and one should be careful when working with linguistic data, as even the apparently nonthreatening symbol r is a bottomless trap for the unwary.