For example, in normal conversation, in words like engaged or encouraged, I notice that the d sounds really weak (almost unvoiced), especially when it comes before another consonant--as in We're very, very encouraged with how things are going.

Does this really happen in normal conversation, or is it just my impression?

  • Here are some relevant questions on the English Language and Usage Stack Exchange that you might want to take a look at: the elision of alveolar plosives, pronunciation of t after c and before l – brass tacks Oct 28 '15 at 3:58
  • Wells 2008: the elision of d at the end of the word (before a consonant at the beginning the next word) is more common in ld and nd and less common in bd, gd, dʒd, vd, ðd, zd, md, ŋd. – Alex B. Dec 27 '15 at 17:31
  • see Cruttenden 2014, pp. 314-315 for more details. He has very interesting examples, esp. "He walked back" or "They seemed glad." – Alex B. Dec 27 '15 at 17:40

There is a debate over whether the p/b distinction is based on voicing or aspiration, owing to the fact that the ostensively voiced stops are rather different from those of French or Hindi. Between voiced segments (especially vowels) they are voiced, but word-initially and finally they may be voiceless unaspirated stops and not voiced (this is speaker-dependent and also dependent on place). There is a complex set of factors that govern whether there will be vocal fold vibration. In "encouraged", the factors favoring devoicing are very high -- following a long period of obstruent closure, syllable final position, no following voiced segment. So it is quite likely that vocal fold vibration has been extinguished in this context. This is a non-neutralizing process, so even if vocal fold vibration has ceased in the final stop of "edged", it is still distinct from "etched". You would have to look at the phonetic details of a particular example recording to say what difference, if any, could be localized to [d] in this pair.

Not the answer you're looking for? Browse other questions tagged or ask your own question.