Guttural deletion in loanwords 1

Carole Paradis a1 and Darlene LaCharité a1
a1 Laval University


Among 4,499 segmental malformations found in English loanwords in three large corpora of French, the laryngeal /h/ is the only segment that is never adapted, i.e. replaced by another segment. We suggest that the systematic deletion of /h/ in French follows from the fact that, phonologically, French, like Portuguese and Italian, does not employ the Pharyngeal node, the articulator that characterises gutturals. This prevents English /h/ from being handled phonologically (deleted or substituted) in those languages. The non-availability of the Pharyngeal node also explains systematic deletion of the pharyngeal and laryngeal gutturals in Arabic loanwords in French. In contrast, English /h/ is adapted by languages employing the Pharyngeal node phonologically, such as Spanish, Bulgarian, Catalan, Mandarin Chinese, Greek and Russian. Likewise, the availability of the Pharyngeal node in Fula and English allows the adaptation of Arabic pharyngeal and laryngeal gutturals in Fula, and non-glottal gutturals in English.


1 We would like to thank the editors of Phonology and the anonymous referees for their helpful comments in connection with the preparation of this article. We also benefited from the comments and questions of Ahmed Alioua, Jacques Durand, Ali Idrissi, Michael Kenstowicz, Pierre Martin, Jean-François Prunet and Charles Ulrich on previous versions of this article. Ahmed Alioua, a phonetician from Safi (Morocco), and Nadir Kerris, from Setif (Algeria), checked the Arabic pronunciation of our Arabic loans in French, while Abdulhamid H. Gadoua, a PhD student in phonology who was born in Libya, verified the Arabic pronunciation of our Arabic loans in English. We are particularly indebted to Frédérick Brault, one of our research assistants, for having collected the loanwords of the list appended to this article, checked them with native speaker consultants of European Spanish, Bulgarian, Catalan, Mandarin Chinese, Greek, Russian, Italian and Portuguese, and for having taped, transcribed and computerised them. F. Brault also presented with C. Paradis a preliminary French version of this article at the annual meeting of ACFAS in 1998, and published that version with C. Paradis and D. LaCharité in 1999. We are also extremely grateful to Fatim El Fenne for her transcriptions of Moroccan and Classical Arabic, and for the time she devoted to the collection of Arabic borrowings in French. We are indebted to Abdourahmane Sakho for his help with the pronunciation of Arabic borrowings in Mauritanian Fula and to our colleagues Fernão Perestrello and Alexander Sadetsky for their help with Portuguese and Russian, respectively. Thank you also to our MA student Élyse Bolduc for having processed her English loans in Mexican Spanish earlier than expected in order to provide us with statistics rapidly. More generally, we would like to thank all of our consultants and research assistants for their precious collaboration, and the people who kindly answered our questions on the Linguist List. Naturally, we are solely responsible for interpretations and possible errors. Research for this paper was made possible by SSHRCC grants #410-97-1446 and FCAR #98-ER-2305 to C. Paradis, SSHRCC grant #410-2000-0337 to C. Paradis and D. LaCharitéand by a grant from the Faculté des Lettres at Laval University to D. LaCharité.