View Message

It should also be mentioned that certain consonant pairs do not occur in speech, as each cluster must be either consistently "voiced" or "unvoiced/voiceless". If one part of the pair must be unvoiced, the normally voiced consonant will be unvoiced to it's associated allophone/phoneme. Thus a word which grammatically would end with the voiced |d| will instead be pronounced as if voiceless |t| if the preceding consonant is also voiceless (e.g. slept, slapped, wept, left, sipped, dipped, blessed), and retained if the the preceding consonant is voiced or there is an intervening voiced vowel (e.g. webbed, waved, loved, subbed, dubbed and blazed). This also occurs the other way around - a normally voiced consonant may be unvoiced if the following consonant is voiceless, and normally unvoiced consonants may be voiced (house v. housed). Grammatically the former is rare, but is frequent in compound names such as Tipton - in this case neither the i and o are long enough to force the consonant pair to be voiced (if they were it may have become Tibbedon), so if the if the eponymous individual was Tibba, it would be forced to to Tipton anyway (the short vowels i and a explain why a Saxon Tibba would soon become Tippa as well - the same forces are at work in names such as Padraic, Patrick, Patty and Paddy).