N°020_Vol.4_14

Introduction: Borrowing is among the most ordinary facts of linguistic life. No speech community is sealed off from its neighbours, and sooner or later every language takes in words that were shaped elsewhere: by conquest and trade in earlier centuries, by screens and markets today. For the phonologist, however, the interesting moment is not the act of borrowing but what follows it. A foreign word seldom settles into its host untouched. It is trimmed, re-tuned and bent into shape until it obeys the sound pattern of the language that receives it. This re-tuning is what we shall call phonological adaptation, and it is the subject of the present article. Loanwords are, for the student of sound systems, an unusually clear laboratory: in them we watch a grammar operate on material it did not itself produce. English is a convenient place to begin, because it sits on both sides of the exchange. Across its history it has absorbed enormous quantities of Latin, French and Scandinavian vocabulary, and in doing so it has built up a varied repertoire of adaptation habits. Yet English is today also the great donor language: its words travel into Japanese, Korean, French, Akan and Yorùbá, where grammars quite unlike its own take them in hand. A single language therefore lets us observe adaptation from both ends at once: as the language that adapts, and as the language whose words are adapted by others.

This double position frames the question that organises our study. If each receiving language remakes a borrowed word in its own image, what exactly fixes the result? Is the adapted form the work of the phonological grammar, of perception, or of some traffic between the two? Around this question we set three hypotheses. The first (H1) is that the shape of an adapted loanword is determined in the first instance by the markedness ranking of the receiving grammar, and only secondarily by faithfulness to the source form. The second (H2) is that languages share a single preference scale among repair strategies, one on which deletion is the universal last resort: segmental repair is ordered substitution before importation before deletion, and prosodic repair likewise prefers epenthesis, which keeps every segment, over deletion. What varies across languages is not this ordering but its realisation, set by language-particular properties such as vowel harmony, tone and the syllable template. The third (H3) is that neither a purely perceptual account nor a purely grammatical one is sufficient on its own: an adequate model has to let perception and grammar interact.

Our objectives follow directly. We aim, first, to gather within one comparative frame the segmental and prosodic adaptations attested in ten genetically and typologically diverse languages; second, to weigh the three hypotheses against that material; and third, to ask which of the competing theoretical accounts accommodates it best. The method is descriptive and comparative; this is a synthesis of the published descriptive record rather than a report of new experimental data, so the three hypotheses function as organising theses to be weighed against that comparative record rather than as predictions tested against fresh instrumental findings. We draw on the published adaptation literature, including recent descriptions of West African languages, and we cast the recurring generalisations in the vocabulary of Optimality Theory and of Feature Geometry. These tools are not chosen at random. Optimality Theory gives us a transparent way of stating that the grammar of the receiving language outranks fidelity to the source, precisely the claim of H1, while Feature Geometry makes explicit the notion of segmental “closeness” on which every account of substitution silently depends. Haugen, Van Coetsem and Silverman, for their part, supply the three classical vantage points (typological, directional and perceptual) from which the data can be read.

The stakes are clear. As C. Paradis & D. LaCharité (2011) emphasise, a borrowing language does not simply reproduce the phonological structure of the source; it filters that material through the constraints of its own grammar, replacing impossible segments, adjusting syllable structure and reassigning prosody in ways that lay bare the internal logic of the receiving system. Words such as élan, nuance, credit and desk, passing between French and English, undergo precisely this kind of obligatory adaptation. The article is organised around these commitments. Section 1 sets out the three frameworks on which we build. Section 2 turns to English as a borrower, and Section 3 to the prosodic side of adaptation. Section 4 reverses the direction and follows English words into European, African and Asian languages. Section 5 weighs the competing theories against one another, after which the conclusion returns to the three hypotheses and asks whether the evidence has borne them out.