<< Most Pāli grammars (Collins, Geiger, Duroiselle) produced in the west
give the form of the possessive suffixes as -vant/-mant. But Narada's Intro,
and Buddhadatta's dictionary follow the Pāli grammars in making the stem
form -vantu/-mantu.
Since the 'u' plays no role in the declensions I'm interested to know why it
is there, especially since western scholars silently dropped it in most
cases (except for a very brief note in Warder's intro p.252, n.1).
Can anyone shed light on this form? >>
The 'u' is called an anubandha (indicatory letter) which is dropped in
word-formation. The traditional grammars, both Pali and Sanskrit, when
giving a root or suffix frequently show them with an attached anubandha
letter (sometimes more than one). They have a purpose, e.g. an '.n'
indicates that a vowel in the root undergoes vuddhi (increase).
Compare Pali mantu/vantu with their Pa.ninian Sanskrit counterparts:
matup/vatup.
I don't know the purpose of having a 'u' at the end of mantu or vantu
instead of some other vowel. That would require some investigation.
Explanations of the anubandhas can be found in S.C. Vasu's English
translation of Panini's A.s.taadhyaayii.