In this chapter I will attempt to use these rare examples of declension to get a bit more systematic insight into the grammatical system of Minoan language. This is a mammoth task, given the scant evidence, and will likely not be completed until all Minoan texts (including the (in)famous Phaistos Disc) are fully deciphered. So I warn all readers in advance: all the theories to be discussed here have an extremely high uncertainty factor.
To begin with, let us consider a few examples drawn from the Linear A tablets. The following groups contain words with (supposedly) related stems. I am also going to give a (highly tentative) reading of the words, to help imagine how these suffixes looked like in the real language.
Our first example will be the stem *PI-TA-K-? (perhaps *pinthake?). We can see two different derivatives of the same stem on the Haghia Triada tablets:
HT21: PI-TA-KA-SE (*pinthak-ase?)
HT87: PI-TA-KE-SI (*pinthake-si?)
What is really interesting, are not the endings themselves, but rather, the way they fit onto the stem. Throughout all the tablets, we can find many words ending with either -SI or -SE. Their stem-vowels follow a very similar distribution, with -A- being the most common, followed by -I-, but -E- is really that rare. But how do we explain the discordance of stem-vowels found in our example? It cannot be just some 'scribal preference' or 'dialectal variation' (both examples come from the same place). One cannot shake the feeling that this is a regular grammatical change.
There are a handful of good examples on the tablets to reinforce the feeling that the ending seen on PI-TA-KA-SE is not just -SE. It is -A-SE. Throughout the corpus of inscriptions, there are many other similarly-suffixed words (mostly names), i.e. A-SE, DI-DI-KA-SE, DU-RE-ZA-SE, etc. And one should not forget about the fairly similar Eteo-Cypriot declensional ending -O-SE (i.e. A-RI-SI-TO-NO-SE = *Ariston-ose = "from/of/to Ariston" on the Amathus stele).
In some of these words (our example was such one), the -A- vowel (-O- in Eteo-Cypriot) seems to 'intrude' into the stem, replacing its original ending, in case if it ended with a different vowel. In this sense, we may call many Minoan suffixes as 'intrusive', as they do not simply add on to the stem-words (our example was very likely that of a noun), but instead replace its base ending with their own vowel. What we see on the above example-pair is an intrusive suffix beside a non-intrusive one. -A-SE is probably intrusive because this declensional ending starts with a vowel. On the other hand, -SI does not seem to change the stems (at least in our case), so it does not necessaritly use an initial vowel, such as -A-.
What do these suffixes mean? Since we have no useful etymological counterpart of PI-TA-KE, it is hard to tell what it meant. Chances are good that it was a name. The Eteo-Cypriot endings related to -A-SE occur on names, and probably mean either ablative (less probable), dative or genitive cases. If -A-SE, let us say, is a genitive case, then -SI could be a related ablative case ending. But at this point this is a mere game with thoughts. We will need more solid evidence before we can identify the cases with more-or-less certainty.
But let us continue our search for other examples of declension. This time we will tackle the very common -A-RE ending, to see if we can see at least one example of a word loosing this ending. It may not be the best linguistic example ever, but the stem A-R?-N-? (*arne?) will do for this time. We have the following three examples:
HT1: A-RA-NA-RE (*arn-ale?)
KNZf13: A-RE-NE-SI- (*arne-si?)
HT25: A-RI-NI-TA (*arn-intha?)
If the division of signs on the golden ring of KNZf13 is correct, we have a word with the previously-seen -SI ending. On the counter-example it stands with the (supposedly intrusive) -A-RE ending. This ending is incredibly common. A good percentage of all words use this very case-ending. Since it dominates on tablets listing personal names, it is credible that this is either a typical genitive or ablative case ending of Minoan words (dative makes less sense, given the context: taxation).
The last parallel to this stem shows another typical case-ending: *-intha. From the context, it is probably an adjective-forming suffix. We can reconstruct it with more-or less certainty since it is the same ending so characteristic of "Pre-Greek" place-names found around the Aegean sea. Korinthos is just one of the many cities that bear such a Minoan-era name. We can see from our example that this ending is - again - intrusive: The -i- vowel almost always precedes the -nth- cluster in these Pre-Greek toponyms as well, still corresponding to ancient Minoan grammatical rules after the Cretan civilization disappeared into oblivion some 3000 years ago.