CONTINU!TY AND CHANGE

1

To the memory of Kudzi

:l

Studies in African Archaeology 13

CONTINUITY AND CHANGE
An archaeologica/ study af farming communities in northern Zimbabwe
AD
by Gilbert Pwiti

500-1700

Uppsala 1996

Doctoral thesis at Uppsala University 1996 Printed with the aid of grants from Uppsala U niversity and the Swedish Intemational Development Cooperation Agency (Sida)/Department for Research Cooperation (SAREC)

ABSTRACT

Pwiti, G. 1996. Continuity and Change: an archaeological study of fanning communities in northern Zimbabwe AD 500-1700. Studies in African Archaeology 13, Uppsala, 180 pp, 43 figs, 21 plates ISSN 0284-5040, ISBN 91-506-1159-3. Ulls thesis addresses two main questions about the iron-using farming communities in northem Zimbabwe: when did they become established and how - - :::
____(

'

\'

,_

,,,
___

Eo

z~----

Figure 25. Mutota site plan.

81

\

:

"' 0 0

"'

\j

I

~-· -~ o

~----··

"'

\
\
N

//
Figure 26. Matusadona site plan.

y
0

I

affinities with that recovered from both Great Zimbabwe and Khami. Copper objects, the most common being wound bangles and copper beads were also recovered. In addition, ivory bangles were recovered (Plate 20). Glass beads A total of 120 glass beads of different colours were recovered mostly from the midden. Their classification is summarised in Table 1la. Despite the occurrence of fragments of moulded dhaka showing house remains on the surface and in the test pits, no actual house floors were found. One charcoal sample recovered from Test Pit I, and another from Test Pit Il were collected and submitted for C14 dating. The results are presented in Table 19. Kasekete pottery: description, classification and analysis Initially all the 4766 excavated sherds were categorised into vessel parts of rim, nyck, shoulder, body and base, or combinations of these. These produced 2089, mostly graphited, classifiable sherds although
Figure 27a. Kasekete: Test Pit I section (south).

§

UGHT BROWN GREYISH As'.~y

tJ
~

UGHT GREY ASHY

YELLOWISH BROWN ASHY

ELJ
(fil

Wl-11TE GREYISH ASHY

YELLOWASH

[J

STERILE

IOOCM

82

§l

BROWN HUMUS/STONES YELLOWISH GREY REODISH GREY

El

DARK BROWN

HUMUS

!i)

D l!!l

REDDISH GREY
REDDISH GRAVELLY
STERILE

!lilll

ILl

WHIT\SH GREY
REDOISH GRAVEL STERILE STONES
0

EJ

IJII

!LI
111
0 lOOcm l 06 cm

Figure 27b. Kasekete: Test Pit Il section (south). Figure 27d. Kasekete: the Test Pit VIa section against the southern wall of the rectan ular enclosure.

Figure 27c. Kasekete: Test Pit fil section (south).

---walling above surface surface ---brownish humus braken pot

- - - stone walling

::···-

.-:

- - - pottery and ash - - - sterile

0

1M

83

body sherds and bases presented serious problems of identification (see Table 13). In this case only vessels that could be clearly assigned the identity of 'body' or 'base' on the basis of profile or contour and sometimes possessing some graphite burnishing and or decoration, were considered. This left a total of %77 (72 graphited) body sherds which were mostly plain and could not easily be assigned the correct vessel part. Decoration was primarily categorised according to dominant technique, which was then cross tabulated to determine the most common technique/motif combinations. Owing to the high rate of sherd fragmentation, no multidimensional typologies as suggested by Huffrnan (1989a) were performed as it was difficult to determine decoration placement, or full motif layout. For the present analysis technique/motif combinations are adequate since through their combination of different decorative traits, they clearly show the character of an assemblage. On the basis of defined vessel parts, eight shape profiles were identified. These are:
Bowls (see Fig. 28a) 1. Slightly constricted bowls 2. Open hemispherical bowls 3. Necked bowls Pots (see Fig. 28b) 4. Pots with concave rims 5. Pots with insloping rims. 6. Pots with short vertical necks 7. Pots with tall vertical neck/rims 8. Wide mouthed, shouldered pots with rim diameter equal to or slightly less than the maximum diameter. Owing to the degree of vessel fragmentation, only a few illustrations are presented here. A fuller presentation will appear in a separate research report. Table llh provides a summary ofthe vessel shape profiles. On the basis of the excavated material, there is no differentiation of vessel shapelprofile by layer. Mast of the material came from the upper levels of Test Pits I and VI while same test pits yielded little (see Table 14). Kasekete pottery is therefore composed of a single homogeneous assemblage. The composition of the assemblage, coupled with the continuous occupation strongly suggests that Kasekete was a single phase site. All the vessels recovered are generally well polished with a grey to brownish grey colour on both exterior and interior surfaces, including the core. Some are graphite burnished on the exterior surface and this application extends to cover the interior upper parts of the rim. Thickened lipforms are extremely rare. In

one case where it has been identified, it seems to have been a result of defining the neck rather. than a deliberate act of adding more clay. Lip profiles are simply rounded or tapered, with their final appearance determined by rimlneck morphology. Decoration is extremely rare. An examination of the decorated sherds from surface contexts would have provided a more reliable indication. of the character of the assemblage at Kasekete but most of the material was surface collected by previous investigators. Only a few sherds (including a complete pot) are at the Rarare Museum of Human Sciences, and these seem to indicate that the use of applied designs involving graphite bumishing and red ochre was central to the decoration themes on the site. The dominant decoration techniques are: comb stamping, incision and painting/polishing. Punctates are limited to three sherds while bosses have been recorded on one sherd. These techniques produce pattem elements ranging from single horiwntal lines to more complex geometrical triangular and sometimes rectangular designs (Fig. 28c). The decoration motif can be described within the defined technique categories mention above. Moulding is a morphological character with a decorative effect. One sherd had two bosses applied on the shoulder/body region (see Fig. 28b). From Table 12, it is clear that applied techniques combine with all the other techniques, and this clearly shows how central this was to th_e decoration of the pottery found at Kasekete. Discussion The Kasekete assemblage broadly campares with excavated material from other Great Zimbabwe Tradition sites in northern Zimbabwe like Nhunguza, Ruanga (Garlake 1973b) and Zvongombe (Zishiri 1990) in terms of vessel shape profile. There are, however, notable differences in rim morphology and decoration. While the dominant shape at Nhunguza and Ruanga was the vessel with spherical body, short neck, slightly concave or straight and vertical or slightly inwardly inclined profile, Kasekete potters preferred to make vessels with insloping rims. This is also the trend at the two stone enclosures on Zvongombe Hill in Centenary, but there is a difference in the emphasis on the lip, which is externally or internally 'rol!ed' for Zishiri's (1990) Type 3 vessels which, like his Type 1 have an insloping rim. Five vessel shape profiles have been defined at Nhunguza and these match profiles 1 and 2 and 4 to

84

Table lla. Glass beads from Kasekete Size Medium Small Medium Medium
Colour Light blue Light blue Dark blue Yellow Light blue

Table l lb. Summary of vessel shape profiles at Kasekete
Total

T-ype
Cylinder Oblate Ob!ate Oblate Oblate Oblate Barret Oblate Barrel Oblate Round tabular Barrel Cylinder Cylinder Barrel Oblate Oblate

6 4

Medium
Small Medium Medium

Medium Medium Medium
. Small Small

easily isolate sherds of the two traditions on the basis of finish and decoration. A common feature at all the Green Small sites in question is the rarity of bowls. This absence Green Medium seems to be a typical feature of the Great Zimbabwe Light blue Barrel 1 Medium Tradition assemblages. There is, however, noticeable White Oblate Medium I change at the 16th/ 17th century site of Baranda in the Dark blue Cylinder 2 Medium Cylinder disk Dark blue 2 Small Chesa area of Mt. Darwin where the inhabitants who Yellow Oblate l Small were extensively trading with the coast, started to place Total 120 greater emphasis on the production of open hemiSize classification: spherical, constricted and shouldered/necked bowls small = 3 and