Phonological opacity

In phonology, opacity is when a phonological rule that exists in a given language appears to be contradicted by the surface structure (i.e., actual pronunciation) of words in the language. The term was first defined by Kiparsky in the following way:

A phonological rule P, ${\displaystyle A\rightarrow B/C{\underline {\quad }}D}$, is opaque if one of the following surface structures exists:

• instance of A in the ${\displaystyle C{\underline {\quad }}D}$ environment;
• instance of B created by P in an environment other than ${\displaystyle C{\underline {\quad }}D}$;

A common example is the interaction of the flapping of /t/ and the raising of /aɪ/ in Canadian English and other dialects. Before voiceless consonants such as [t], the diphthong /aɪ/ is raised to sound more like [ʌɪ], so the word write is pronounced [rʌɪt]. In some contexts between vowels, /t/ is replaced by the (voiced) flap [ɾ], so (for example) patting is pronounced [pæɾɪŋ], similar to padding. In words like writing, where /aɪ/ is followed by /t/ between vowels, both rules apply and the result is [rʌɪɾɪŋ]—with /aɪ/ raised to [ʌɪ] and /t/ flapped to [ɾ]. This output therefore has [ʌɪ] immediately followed by a voiced consonant, even though the rule that produces [ʌɪ] only applies before voiceless consonants. The interaction is thus opaque: [ʌɪ] is present on the surface in an environment that differs from the environment in which the rule that creates it applies.