Many languages exhibit voiceless vowels, by which is meant the pronunciation of a vowel sound in certain contexts without the vibration of the vocal bands. Voicelessness typically precedes vowel loss, as in English lone from alone or round from around. This sort of phenomenon can be observed in the speech of Barack Obama, who routinely either unvoices or drops the initial vowel in America(n), as does the (cloying) radio host Ira Glass (“This American Life”), illustrating what is called APHESIS, defined as the loss of a short unaccented vowel at the beginning of a word.
The loss of a vowel and/or an adjacent consonant can also occur in the middle or end of a word, in which case it is called syncope (medial syllables, as in bos’n for boatswain) or apocope (final syllables, as in sing < Old English singan). Such phonetic processes are first observed historically in colloquial or allegro tempo varieties of speech (for instance, the so-called loss of the “jers” [= supershort vowels] in medieval Slavic) and are then generalized to all styles regardless of tempo. In Japanese the high vowels i and u are regularly syncopated between voiceless consonants in all styles unless emphasis is called for, in which case they can be reinstated in lento tempo. French routinely syncopates medial vowels in neutral (elliptical) speech (cf. maintenant, etc.), reinstating them when called for in explicit style.
What vowel loss illustrates is the INDEXICAL FUNCTION of contextual variation in language. Aphesis and syncope are always tied to specific phonetic contexts, and they are thus SIGNS––INDEXES, to be precise––of both the value of the vowel involved, on one hand, and the value of the consonants in the segmental context, on the other. This semiotic grounding of the phenomena at issue insures the solidarity between rules and contexts, absent which the phonology of a language––like all the components of grammar––would not be the coherent system that it is.