Target Lg noun subclass with its own form of a suffix

Question

In the source language Mato, Proper Nouns take an associative suffix -am in the same way any other noun does:

However, in the target language Uma where the normal Associative suffix is -an, Proper Nouns seem to take their own form of the suffix that does not follow phonological rules, but is assigned according to its subclass:

Yesu Nasaret-gano (*Yesu Nasaret-ian) "Jesus of Nazareth"
Josep Erimatea-nggano (*Josep Erimetea-n) "Joseph of Arimathea"

With limited data, it appears the Associative suffix for proper nouns is -gano / [C]_ and -nggano / [V]_

How do I handle this in the FLExTrans process? My inclination is to create a target lg template for the Noun subclass Proper Noun and populate it with the ASS5 suffix -gano (which would be a separate entry in the lexicon, not a separate sense since I have no way to distinguish the suffix's form in the Allomorph section), then write a choose statement rule under the Noun rules that if its a Proper Noun and the Associative slot contains "ASS", then set it to "ASS5".

Will this work? Is there a different way to approach this?

Answer 1

Yes, that solution would work well. I think when you add the template for Proper Noun and add a new slot, you could just add the same Associative slot you have for nouns and add -gano ASS5 to that. I don't think it buys you anything to create a new slot for it. Make sure the -gano suffix is something that attaches to nprop.

Your rule idea is good and seems to work for me.

One thing to keep in mind. There will be two templates that will be allowed to apply to Proper Nouns one at the Proper Noun level and one at the Noun level. This means a word in an Uma text like Nasaretan will be allowed to parse because it is applying the top level template. You may not need to worry about checking the spelling of Uma words with the parser, but if you want this to be disallowed, you can add a custom inflection feature to all your proper nouns indicating 'properness' and a feature to -an of 'unproperness' and it would disallow the two from co-occurring. I tested this.