The sierra was reportedly occupied by peoples speaking Barbacoan, Jivaroan and unclassified languages by the mid-16th century; Quechua seems to have been the language of an urban elite at that time.
Based on the extant descriptive literature on Quechuan, Barbacoan and Jivaroan, and concentrating on the loss and restructuring of EQ morphology, this study concludes that there is still a substantial amount of work to be done in order to render Muysken's intuitively attractive and possibly even correct case stronger.
(4) The non-Quechua languages chosen--because of both data availability and spatial proximity to EQ--are Awa Pit and Tsafiki (Barbacoan) on the one hand and Aguaruna and Shuar (Jivaroan) on the other.
Jivaroan languages, by contrast, has argument markers closer in function to the Quechuan ones--and note that both Aguaruna and Shuar actually have portmanteau morphemes for 1[left right arrow]2 interactions:
The loss of clusivity in EQ, on the other hand, is more plausibly attributed to contact with clusivity-less Barbacoan and Jivaroan, even though a contact-independent explanation is certainly also possible.
As far as I can judge from the extant descriptions, Barbacoan languages do not have benefactive applicative markers, but those found in Jivaroan (particularly in Aguaruna), are well-behaved (i.e., they are not valency-neutral like the present-day EQ morpheme), i.e.
Jivaroan languages, by contrast, have switch-reference forms that do mark person in such clauses.
Limiting myself here to the second hypothesis, it is clear that both Barbacoan and Jivaroan languages have morphologically simple patterns with respect to subordinate clauses: adverbial and purposes clauses show switch-reference forms in all languages (like EQ), but these do not mark person in Barbacoan, even in different-subject forms (like EQ) and do mark person in Jivaroan, even in same-subject forms.
I was not able to see any other connection between these developments, especially the latter two listed above, and possible Jivaroan or Barbacoan models.
There does not seem to be a direct motivating or driving force in either Barbacoan or Jivaroan, let alone in Spanish, for such developments.