[Note: this transcription was produced by an automatic OCR engine]

xxxvi INTRODUCTION
same word was apparently pronounced slightly differently by
different informants. Thus there is a guttural uvular “ 1 â€ù which
tends towards a strong guttural " gh " sound ; the former Deacon
wrote as " rh " the latter as " gh â€ù, and what are undoubtedly
the same words sometimes appear containing the former symbol,
sometimes the latter, but not infrequently the diacritical mark
is omitted entirely. Again, in the dialect of Seniang there is a
“ Z2 â€ù sound, in the forming of which the lips are made to vibrate :
this is sometimes written " 5", sometimes " br ", and sometimes
“ bw â€ù. For example, we ï¬Ånd a word written as nimliunang,
nimbrunong, and nimbwunong, and here again Deacon sometimes
omitted the diacritical mark so that we also get nimbunong.
The same inconsistency is apparent in the transcription of the
typical Melanesian "ml2â€ù sound; sometimes it is written as
"mb â€ù, sometimes as " b ". Where the evidence appears to
warrant it, I have tried to introduce some consistency into the
spelling, but in view of the numerous different dialects in this
island it has often seemed safer to sacriï¬Åce what may after all be
a false consistency for the sake of possible greater accuracy, and
to follow the spelling given by Deacon. In another respect also
I have followed Deacon's example rather than attempt to adhere
to what would perhaps have been from a linguistic point of
view more strictly accurate. In Malekulan the deï¬Ånite article
is represented by the preï¬Åx ne-, ml, no-, no», nu-, or, often before
a vowel, n-. Thus we have Mai “ the tree ", nivaal " the war â€ù,
nembet “the bread-fruit", noumbou or nu/mlmu “the pole".
Sometimes it is clear that the vowel following the n~ is a part
of the article, but often it is not possible to tell whether this is
so or whether it forms part of the word root. Further, it seems
that two homophonous words are sometimes distinguished by
the different forms which the preï¬Åxed deï¬Ånite article takes, as,
for instance, nalang " the wind and nilang, which is the name
apparently of a variety of pudding cooked on special occasions.
There appear, however, to be a very great number of homophones
or homonyms in Malekulan and it is possible that the distinction
according to the form of the article, such as that given above,
is fortuitous, and that the wind too may sometimes be pronounced
mlang. Be that as it may, the evidence suggests that the use
of ne-, m’-, etc., is not strictly comparable to our use of the word