Page 89 - KBHA Bulletin 15

86

EXCAVATIONS IN KALK BAY AND FISH HOEK ROCK SHELTERS AND THE

ORIGINS OF OUR SPECIES: A NARRATIVE OF ON-GOING WORK

John Parkington and Cedric Poggenpoel

The early days: the 1920s–1930s

The late 1920s were eventful years in the development of a terminology for the long history
of the South African stone age. Just after the middle of this decade Stapleton and Hewitt

excavated in a small cave in a valley near Grahamstown and discovered a hitherto
undescribed assemblage of finely made stone artefacts that they named Howiesons Poort,

after the valley (1927, 1928). By the end of the decade John Goodwin and Peter van Riet
Lowe (1929) had produced their epic classification of Early, Middle and Later Stone Age

assemblages from the subcontinent, attempting a chronological placement of the rapidly

accumulating evidence. (See Explanatory Box.) In between, Victor and Bertie Peers had
started to excavate a series of rock shelters and caves in the Trappies Kop and Skildergat Kop

kranzes near their home at Fish Hoek on the Cape Peninsula. The link between these sites,

these events, these people and the current global interest in the origins of our species is
explored here.

In these early days, before any kind of absolute dating, stratigraphy and typology were the

twin keys to understanding the age, associations and likely significance of excavated remains.
John Hewitt was a biologist at the Albany Museum in Grahamstown, a scientist who

understood the need for attention to detail and who excavated as carefully as anyone in his

day. His methods were crude but no more so than those used at the classic French excavations
at Dordogne sites in the early twentieth century. By today’s standards, however, all of these

left much to be desired: ‘pick and shovel’ archaeology, like that also used by the Peers’, is
now a thing of the past. Hewitt’s attention was drawn to what he considered the most defining

artefacts from the assemblages, a series of curved blunted or backed blades that he referred to
as ‘crescent-shaped’, averaging about 3cm long. It was entirely usual in those days to select a

‘type tool’ with which to characterise an assemblage, to make comparisons with other

assemblages and through which to try to make an age assessment. It is significant to this story