[YG Conlang Archives] > [saweli group] > messages [Date Index] [Thread Index] >

[Date Prev] [Date Next] [Thread Prev] [Thread Next]

change to argument structure suffixes

This is an explanation of a major change in the argument structure suffixes and how they're used.  They are still compositional, but I have added some new components, and the meaning of the resulting suffix has changed.  Instead of the suffix representing the new argument structure of a word, it now represents only the change from the default.  In some cases the form of the suffix, including the new "xo/zo", will be unchanged.
Here is a table of the meanings of the components:

s/d                      Focus          A|P
x=>s                w=>+F          a=>P
z=>d                y=>-F          e=>AP
v|0=no chng    0=no chng.        i=>A/P
                                                   o=no chng.

Examples: xa, xe, xi, xo, za, ze, zi, zo, va, ve, vi, wa, we, wi, wo, ya, ye, yi, yo, xwa, xwe, xwi, xwo, xya, xye, xyi, xyo, zya, zye, zyi, zyo.

Each column now has a value for 'no change': 'v|0' for s/d, '0' for F, and 'o' for A|P.  When only the A|P changes (and not s/d or focus) then 'v' is the no-change value (so that there is a consonant).  If 'x', 'z', 'w', or 'y' are used, then 'v' is not used.  This means that the possible suffix "vo" will never occur, since it would mean there are no changes.  I believe this is the only case of duality or context dependency in the morphology.
'Y' can follow 'x', which means that great care must be exercised to distinguish between "xa" and "xya", for example.  Perhaps inserting a short 'i' between "x" and "y" will solve this (in the spoken form only).  Thus "xya" might be pronounced "xiya", while still being spelled "xya".

The focus suffixes are now "wo" (add F) and "yo" (remove F), instead of "we" and "wo", resp.

I haven't determined how these new suffixes interact with the verb endings, which already show the number of arguments and the s/d of the classifier.
Classifiers "za/za'a", "zo", and "zu" don't have standard argument structures, so their suffixes haven't changed.  The new interpretation of these suffixes with these is now that their defaults are all nulls.

This scheme for argument structure suffixes is relative, while the previous one was absolute.  Before, the classifiers had defaults which were not indicated anywhere in their morphology except by which suffixes they took.  The default was the suffix that was never used with that classifier.  Now, every suffix indicates the change from the default, so more information is given about the default than before.  

It may be that this scheme is too cumbersome or hard to remember to use.  In this sense, it's experimental.  If it proves to be unwieldy, it's always possible to go back to the original absolute scheme, namely, the one based directly on Latejami.

stevo