It seems likely that no polysynthetic language has, within its system of pronominal affixes, a set of contrasts large enough to make all the distinctions in definiteness, indefiniteness, genericity and referentiality which can be made, in free nominal expressions, by various combinations of articles, demonstratives, free pronouns, and adjectives representing discourse status. […] I know of no [polysynthetic] language able to convey more than two values on this semantic dimension by its agreement morphology alone.
Standardized
IF there is polysynthesis, THEN no more than a two-way contrast in definiteness, indefiniteness, genericity, and referentiality is conveyed by the bound pronominal affixes of agreement morphology alone.
[…] in general, polysynthetic languages use argument [af]fixes for a wider range of situations than warrant the use of free [personal, definite] pronouns, but that as one moves from definites to indefinites towards non-referential objects every polysynthetic language will draw the line at some point and use an alternative strategy, such as intransitivization. Once we have better data from a range of polysyntheticlanguages it will be interesting to see whether these cut-offs can be arranged in an implicational hierarchy. (Evans 2002)
[…] in general, polysynthetic languages use argument [af]fixes for a wider range of situations than warrant the use of free [personal, definite] pronouns, but that as one moves from definites to indefinites towards non-referential objects every polysynthetic language will draw the line at some point and use an alternative strategy, such as intransitivization. Once we have better data from a range of polysyntheticlanguages it will be interesting to see whether these cut-offs can be arranged in an implicational hierarchy. (Evans 2002)