I seem to detect a lot more evidence that -de and –pi/si
are endings of a similar grammatical function.
                           The rules that final -u must be preceded by Vowel 3,
and that -o- must not be preceded by Vowel 1, seem, at first
sight, to be confirmed by the Knossos index. An exception is      qa-ra-o
on 58, but the spelling seems peculiar generally. 
                           The rules for genitives in -jo and -o seem OK too.
Some variations between -jo and -o within the same stem must
probably be explained as having a definite function, eg:

wi-ri-ne - o    straight genitive singular qualifying o-po-qo
wi-ri-ni - jo   adjectival nominative, with vowel change. (or genitive plural)
wi-ri-ne - jo  alternative spelling, according to rule whereby Vowel 3 is allowed to be 
                      retained in the ‘adjectival’ form (which also operates with feminines in –ja).                 

The evidence for -qe is poor. 820 is a clear case, but the
others are doubtful.
	I wonder if Pylos ri-ne-ja is the adjectival form from ri-jo,
being a dialect variant of Knossos ri-jo (-no) / ri-jo-ni-ja? 

jo / a-mi-ni-so-de (04-67.1) shows the same grammatical form
as the Pylos jo / i-je-si etc, which I referred to in my last Note. 

	These are very random jottings of a few things that have
come to mind on first looking at Scripta Minoa. Before too long we
must get down to the serious job of implementing the rash undertaking
to provide phonetic values within 2 years. 

                                          Yours, 
                                                        Michael Ventris