Saturday 02 August, 2008
Absolutely nothing to do with
I have recently been consulted by Marķa
Alicia Maldonado in Argentina about a research project she is conducting
on coalescent assimilation in English. By this she means, I think,
the sort of thing that textbooks on English phonetics tell us about and
which affects sequences of [t]+[j] and [d]+[j], changing them into [tʃ]
and [dʒ] respectively. I think she also
means to deal with this as a synchronic phenomenon and as a connected
speech process, so examples like nature and soldier,
presumably deriving from earlier [ˈneɪtjə] and [ˈsəʊldjə]
are excluded, as are synchronic examples such as tune pronounced
as [tʃuːn ] and duty as [ˈdʒuːti].
Many years ago my colleague John Baldwin during a
conversation announced that he did not believe in coalescent
assimilation, or rather that he thought that the term was a misnomer for
what actually happened. Coalescence surely means the fusing of of
two or more things into one, and he maintained that this did not happen.
The more I think about it, in response to Marķa Alicia's questions, the
more I am inclined to believe that he was right.
take the example of didn't you. Assuming that the [t] is
not elided and that it is not replaced by [ʔ], it seems to me that there are
at least three possible pronunciationsː
ju] Here most probably the [t] will be palatalised and the [j] slightly devoiced.
I can readily use this pronunciation in my own speech, though in a rather
ju] Here the [j] is retained,
so there is no coalescence. I think this is my normal, relaxed
pronunciation of this and phrases like it.
is true coalescent assimilation. The two original sounds have
fused into a laminal post-alveolar (aka palato-alaveolar) affricate.
I don't think I personally use this pronunciation.
Admittedly, this is based purely on
intuitions about my own speech and I am not sure how widespread this
pattern is. An objective piece of research on this will be pretty
tricky I think, because the auditory differences are very subtle and, I
guess, so are the acoustic cues which produce them. I have to confess
that for years I taught my students that version 3 was correct, and
questioned those who used version 2 in their transcriptions. Mea