Cosanga (Late Period)

The main characteristics of this type are a light color (orange, pink, light gray, light brown, or brownish yellow), a very even surface that has not been polished (with some exceptions), and a sandy and uniform paste. This type was described by Porras (1975) as Cosanga Ordinario, and my general observations coincide with his. However, there is not much of a distinction between his definition of this type and another type he called Borja Ordinario, and in my samples I did not find sherds that matched the characteristics of the latter that I would not classify as Cosanga. There is no chronological distinction between Cosanga Oridinario and Borja Ordinario according to Porras, although he argues that the latter may predominate in the uppermost part of his sequence (while the rest of the time it simply accompanies Cosanga).

Cosanga is a variable type, but most of the cases conform to the general definition of the type. These apply to Cosanga Fine too, except that this subtype corresponds to very fine ceramics with extremely thin walls, and often painted in white or red. Cosanga unequivocally dominates in the uppermost levels of the stratigraphic sequences and composes the bulk of the ceramics collected in the survey and excavations as a whole. Attempts to break it down into chronologically significant types on the basis of slight variations in surface treatment (it can be polished sometimes), color, thickness or paste (sometimes rough but most of the time uniform) were not successful. With the information at hand, it seems most sensible to argue that this variability has no chronological significance. On the other hand, the differences are not so marked as to merit the definition of other types fundamentally distinct from Cosanga. None of the subtypes initially defined (Fine, Orange/Red, Gray, Polished, Sandy, Sandy Thick, Fine White, "Brick", Grooved), shows a special regional distribution suggestive of variations due to location.

Surface: Very smooth on the exterior but only occasionally polished. The color is more frequently orange, light orange, yellowish brown, light or bluish gray, or light brown; and micaceous elements are always visible in both matte and polished surfaces. Occasionally it can be almost white or very light pink, or red. Very frequently it is orange on the exterior and gray on the interior. Cooking clouds are common too. The finishing technique is different on the exterior, where traces of the smoothing process are rarely visible, while on the interior, bands of striations in various angles are common. These look very similar to those left by certain thinning techniques. When the surface has been polished it can be darker, brown or almost black, with a crackled appearance that seems to be the result of a very thin slip, and shallow grooves resulting from polishing are sometimes visible. It feels a little abrasive when the original surface is not preserved. The surface characteristics of Cosanga Fine are the same, except that it is never polished and sometimes appears to have a thin red slip. Cosanga may have white, red, black, purple or red painting, while Cosanga Fine may have red or white painting or slip. Cosanga was initially divided into Cosanga Gray (later on we realized that Cosanga sherds could just be orange in one side and gray in the other one, or have gray cooking clouds), Cosanga White, Cosanga Red, Cosanga Grooved and Cosanga Polished on the basis of color and surface treatment, but no chronological differences appeared to exist among them. As I familiarized myself with the ceramic materials, it was apparent that these distinctions were not really sharp or worth making (this was especially clear, for example, when some of the different colors appeared in the same sherd), yet to be consistent I continued to sort out sherds into these different groups in the ceramic classification of survey and excavation materials. In the end, the number of sherds that fell into these categories turned out to be minuscule (generally less than 1% in both survey and excavations), which may just indicate variations in the manufacturing technique.

Paste: Sandy, fine, and compact. The color can be the same as the surface, uniform throughout, or else gray in the core or half gray and half orange (the half gray usually corresponds to the interior surface). Some thicker sherds tend to have a more coarse and crumbly paste. These were initially classified as Sandy Thick, but this distinction does not appear to have chronological significance. In Cosanga Fine sherds the paste is extremely fine and feels powdery when one breaks a sherd.

Temper: Very small particles, white and black are the most visible. Particles larger than 1 mm are rare and found mostly in thicker sherds. In Cosanga Fine sherds, temper particles are not even distinguishable from the rest of the paste.

Walls: Between 2 and 5 mm are the most common. Thick sherds, about 8 to 10 mm, are out of the normal range and were initially classified as Sandy Thick. Because this is very rare (it appeared in only 4 stratigraphic tests) it was difficult to establish its chronological position with certainty, but it has a tendency to be associated with Cosanga pottery in survey collections and for now I am treating it as a variety of Cosanga because of its similar appearance. It represents only 0.54% of the sherds collected in the survey, and appears in only 80 collections. Other than that, the walls of Cosanga sherds are consistently thinner when compared to those of Bermejo sherds.

Rims: Everted rims (Illustration) of restricted pots (straight or curved), direct rims of globular unrestricted pots, long and thickened slightly everted rims of jars, direct rims of bowls. Some variations in edges are common (thickened, doubled onto itself, or flattened). Also very common are bases of compoteras (bowls with anular bases of different heights). Rims of dishes are less common. Cosanga Fine rims include only occasionally everted and curved rims of tiny pots, and in general just bowl rims and compotera bases.

Decoration: Painted (negative and positive), modeled, applied, and incised decoration is common. Painted decoration, the most regular, has various designs. It can be black, red, white, or purple, in parallel bands or crossed bands on body sherds or in the interior of everted rims. Sets of short parallel white lines appear frequently on the bodies and rims of bowls and compoteras. Painted designs may include perpendicular bands too, in several angles and with negative circles inside the bands. Others similar to a chess-board pattern and curvilinear designs appear as well. Incised decorations appear in the form of two parallel rows of dots on a rim doubled onto itself, or on a flat direct rim (called Cosanga Ribete Punteado by Porras). Bands of hollow circles are common too (called Cosanga Estampado en Anillos by Porras). Less common are shallow incised parallel bands. Modeled decoration in anthropomorphic designs appears on body sherds that seem to correspond to large pots. Applied decoration includes series or couples of buttons on the rim of compoteras (called Cosanga Bordes con Nudos by Porras). Sometimes these appear to be modeled instead of applied. In Cosanga Fine decoration is restricted to painting. No incised, modeled, or applied decoration patterns are associated with this subtype.

Jijón y Caamaño (1952) as well as Porras (1975) established some chronological distinctions among decoration types (negative preceding positive and incised for Jijón y Caamaño, positive preceding negative and incised for Porras). Lumbreras (1990) also suggests a division of Cosanga (Early, Middle, and Late) based on the same decorative characteristics defined by Jijón y Caamaño and Porras, contradicting the chronological order that they proposed (Lumbreras argues that incised decoration and negative painting, not positive painting and slip, serve as chronological markers to characterize the early and late manifestation of Cosanga respectively). In contrast, Athens (1980) and Schoenfelder (1981) argue, agreeing with Jijón y Caamaño and Porras, that changes in the type of painting actually serve as chronological indicators. I did not find any consistent trend in the succession of decoration types in the ceramic analysis of excavated materials, nor did I perceive any difference other than decoration between Cosanga sherds with different types of painting or decoration, but the sample of decorated sherds is small and therefore these observations could change if a larger sample of decorated sherds was analyzed.

Return to "Ceramic Chronology and Stratigraphic Testing" Main Page