NO. 1 CULTURAL DEVELOPMENT IN LATIN AMERICA 109 
with cultures in Bolivia, or across the cordillera from Chile. Either 
appears possible at present, the former because of the affinities be- 
tween cultures such as Tafi II and complexes of the Titicaca Basin, 
and the latter because of the earlier-mentioned relationships between 
Condorhuasi and Molle. 
The culture that brought the Tafi ceramics, brought a number of 
other traits that became widespread during the Early Ceramic Period. 
Among these are a settlement pattern consisting of small houses 
around a central patio. This is found not only throughout northwest- 
ern Argentina, but also in the Puna. It is a pattern shared with cul- 
tures as distant as Chiripa, and one that persists among Tropical 
Forest tribes of today. Associated with this settlement pattern is the 
practice of burial in the house floor or in the central patio, and stone 
wall construction. Another frequently associated trait is the use of 
artificial mounds (Tafi, Alamito, Cienaga, Huasco). Urn burial of 
children in cemeteries (Cienaga), association of grave goods with 
adult burials, stone cists (Condorhuasi in Laguna Blanca), cere- 
monial platforms, stone sculpture (Condorhuasi, Mound Culture, 
etc.) and terraces (Cienaga) are other innovations in this period. 
Comments on the origin and route of introduction of some of these 
elements are of interest. 
Burial urns, particularly for children, are typical of Cienaga, com- 
plexes of the Selvas Occidentales, and the mounds of south-central 
Bolivia. They do not occur in El Molle, and are absent or rare in Con- 
dorhuasi. They are found sporadically in the Puna, where they are 
not grouped in cemeteries as they are in Cienaga, and where they have 
not been dated. This practice could not have come from northern Chile 
or the western Puna, but must have reached northwestern Argentina 
from the north (southern Bolivia) via the altiplano or from the river 
valleys and tropical forest bordering the eastern Andes. 
A similar argument can be made for the origin of grooved axes and 
pipes. Without distinguishing among the latter typologically, it can be 
said that they do not occur in Peru, but are very common in north- 
western Argentina, being present in all the complexes of the Early 
Ceramic Period. In Chile, pipes occur in the Puna region, being par- 
ticularly characteristic of the Molle Culture. This distribution sug- 
gests they must have arrived via one of the routes postulated for the 
introduction of the burial urns. From northwestern Argentina they 
crossed the cordillera into Chile and were incorporated into the Molle 
Culture, an event that must have taken place during Molle I. The dif- 
fusion of the lip plug (tembeta) may have had a similar history, al- 
