740

south in the Bolivian departments of Oruro and Potosi
implies a social/political split at some time in the past.
Such a split would account for the apparent lack of any
gradual shading of features from northern to southern
dialects, although a definite determination in this
respect must await further research in borderline areas

of southern La Paz and northern Oruro departments. Within
an area, existing similar usages may be reinforced by
trade. The intensive marketing and related travel that
occur around Lake Titicaca (Appleby 1976) are obvious

to any visitor. Similar movements associated with trade
also occur in the southern dialect area.

The evidence of a split between northern and
southern dialects does not disprove Torero's theory
(1-2.1) of a probable (gradual?) north to south expansion
of Aymara on the altiplano, since such an expansion might
have taken place before the split, but a simple southward
expansion does not account for the dialectal complexities
revealed by the present study, such as the existence of
(1) certain similar features in noncontiguous areas like
Tarata, Peru and Carangas, Bolivia, (2) features linking
peripheral as distinguished from central dialects or
crossing north-south, center-periphery lines, or (3)
features shared by certain Aymara dialects with contem-
porary Jagqaru and Kawki far to the north. These com-

plexities may be the linguistic reflection of equally