Celtic Mutations

by

Pavel Iosad

• LAST REVIEWED: 21 January 2025
• LAST MODIFIED: 28 February 2017
• DOI: 10.1093/obo/9780199772810-0146

Introduction

The term “initial consonant mutations” refers to a set of alternations that affect word-initial segments (overwhelmingly, consonants) in different phonological and morphosyntactic contexts. For example, in Modern Welsh the citation form of “paper” is papur [’papɨr]. However, the word may also appear as bapur [’bapɨr] (ei bapur “his paper” or Ddarlenais i bapur “I read a paper”); mhapur [ˈm̥ʰapɨr] (fy mhapur “my paper”); or phapur [fapɨr] (ei phapur “her paper”). The form of the word may change because of the presence of certain lexical items (in this case, ei “his,” fy “my,” or ei “her”—in which the mutation can be the sole disambiguation device in certain cases) or in certain syntactic contexts, as when the item is the direct object in a VSO (verb–subject–object) construction with an overt subject. Traditionally, these word-initial mutations are considered together with historically related phenomena that manifest themselves at the left edge of roots rather than whole words (as in Modern Welsh posib “possible” vs. amhosib “impossible,” or Old Irish beir [bʲerʲ] “carries,” ní-beir [nʲiːbʲerʲ] “does not carry” vs. ní-ḃeir [nʲiːvʲerʲ] “does not carry it”). These also are seen with the appearance of certain consonants before word-initial vowels (as in Welsh erthygl “article,” ei herthygl “her article,” or Modern Irish an “the,” ainm “name,” but an t-ainm “the name”), which are often driven by similar considerations. The Celtic languages are particularly well known for their complex and diverse mutation systems that remain an important and pervasive part of their grammar—almost any sentence of any complexity is likely to contain at least a potential context for mutation. The initial mutations in the modern Celtic languages (Welsh, Breton, Irish, and Scottish Gaelic, plus Cornish and Manx—the latter two became quiescent in the modern period, but are now the subject of ongoing revival attempts) do not go back to a common Proto-Celtic system, but instead represent parallel developments in the different branches of the family. Despite the existence of nontrivial parallels in their mutation systems, the Celtic languages also show a surprising diversity of factors driving initial consonant mutations, including phonological, morphological, syntactic, semantic, and lexical properties of both the triggers and targets of the process. The Celtic languages are reasonably well described, and the many descriptions are quite accessible. Perhaps as a consequence of this, Celtic mutations have been the subject of sustained attention from theoretical linguists of many persuasions. Many aspects of their behavior have proved to be challenging puzzles for linguistic theory; therefore, these phenomena remain important testing grounds for theoretical development.

General Overviews

Few comprehensive treatments of mutations in all Celtic languages are available, although the classic paper by Hamp 1951 is pan-Celtic in intent. Pyatt 1997 comes close to providing a comprehensive model; Hannahs 2011 is a general overview focused on perspectives from theoretical phonology. Russell 1995 is an introduction to Celtic linguistics with a chapter devoted to mutations, and several chapters in Andersen 1986 discuss Celtic mutations as a “sandhi” phenomenon. Ball and Müller 1992 is a comprehensive, book-length description of the state of the art at the time regarding numerous facets of the mutation system of Welsh. It is more common to find general works that focus on a particular language while utilizing within-Celtic comparison, or at least proposing approaches that are intended to apply to all Celtic mutations. Willis 1986 is one attempt at such a comprehensive treatment.

• Andersen, Henning, ed. 1986. Sandhi phenomena in the languages of Europe. Trends in Linguistics 33. Berlin and New York: Mouton de Gruyter.

  This handbook of sandhi phenomena contains a number of brief but often useful descriptions of sandhi in Celtic languages, often largely dedicated to mutation.

• Ball, Martin J., and Nicole Müller. 1992. Mutation in Welsh. London and New York: Routledge.

  DOI: 10.4324/9780203192764

  A comprehensive description of the mutation system of modern Welsh, covering the facts of mutation, its diachrony, phonological analyses of the changes themselves, and theoretical accounts of its triggering, sociolinguistics, and acquisition. This book remains indispensable, even if theoretical advances have been made since its publication (this is particularly true of syntax and, to a lesser extent, phonology).

• Ball, Martin J., and Nicole Müller, eds. 2009. The Celtic languages. 2d ed. Routledge Language Family Series. London and New York: Routledge.

  This revised and improved edition of an important compendium of Celtic language descriptions provides the background on individual languages and descriptive data on the mutations themselves.

• Hamp, Eric P. 1951. Morphophonemes of the Keltic mutations. Language 27.3: 230–247.

  DOI: 10.2307/409753

  A classic paper that uses a structuralist framework to account for the triggering of mutation in terms of free-standing “morphophonemic” elements concatenated with mutation targets.

• Hannahs, S. J. 2011. Celtic mutations. In The Blackwell companion to phonology. Vol. 5, Phonology across languages. Edited by Marc van Oostendorp, Colen J. Ewen, Elizabeth Hume, and Keren Rice, 2807–2830. Oxford: Blackwell.

  A general overview of the synchronic data and theoretical approaches to the phonology of mutations.

• Pyatt, Elizabeth J. 1997. An integrated model of the syntax and phonology of Celtic mutation. PhD diss., Harvard Univ.

  A rare example of a comprehensive analysis (focused on Welsh and Breton) that covers both the morphosyntactic aspects of triggering and the phonological rationale of mutation. Also includes some typological comparison.

• Russell, Paul. 1995. An introduction to the Celtic languages. London and New York: Longman.

  A wide-ranging, general introduction to Celtic linguistics that provides both much of the necessary background and a sustained discussion of mutation.

• Willis, Penny. 1986. The initial consonant mutations in Welsh and Breton. Bloomington: Indiana Univ. Linguistics Club.

  An extended comparison of several aspects of Welsh and Breton mutations, with attention to both written and spoken languages and a focus on establishing their place in the grammatical architecture.