Hello all! I suppose this is my first topic I'm posting to the list, I should 
provide a short introduction. I'm a linguist in training (read: soon to be first-
year ling grad student) who's been conlanging since I became interested in 
linguistics, and have been lurking around the list for a while now, but never 
put up the effort to join until recently. Though I've been conlanging since 
before high school, my projects are usually short lived, and seem to flow 
together and change based on my real life linguistic research, so I don't have 
much to show that way. 

Anyway, the real reason for my post. I was thinking about polysynthetic 
languages tonight, and came upon something that I've never quite thought 
deeply on. Namely, is morphological marking of person/number/class features 
of multiple core arguments (subj, obj, S, A, P, whatever your preferred 
theoretical leanings wants to label them) an ::essential:: feature for a 
language to be considered polysynthetic? If one defines polysynthesis as 
simply a high morpheme-per-word ratio, it seems that pronominal morphology   
does not necessarily have to be present. One could conceivably imagine a 
language that had verbs marked for a large number of features (tense, 
aspect, mood/mode, transitivity/valency, evidentiality, etc) but lacking 
pronominal morphology that could still be considered polysynthetic. 

On the other hand, a cluster of features that seem (to me) to accompany 
polysynthesis  (non-configurationality, lack of case marking on NPs) could be 
seen as arising from the presence of pronominal morphology on 
verbs/predicates. There's also the fact that I can't think of any language 
widely labeled 'polysynthetic' that does not include at least some type of 
person marking.

So...does anyone know of natlangs that are considered polysynthetic that do 
not have complex pronominal morphology in their verbs? And has anyone ever 
experimented with making a polysynthetic conlang in which there's no person 
marking?

-- N